Input - Important Systems

From CS2610 Fall 2014
Jump to: navigation, search

Contents

Slides

slides


Readings

Optional

Reading Critique

nro5 (Nathan Ong) 0:13:20 9/21/2014

Review of "CrossY: A Crossing-Based Drawing Application" by Georg Apitz and François Guimbretière As a great extension to Accot's and Zhai's paper, "Beyond Fitts' Law: Models for Trajectory-Based HCI Tasks," Apitz and Guimbretière develop a pen-based interaction program derived from the "Steering Law." Apitz and Guimbretière develop an application that uses "goal crossing" where users, usually with a pen, must cross through a series of goals that replace pointing and clicking at visual elements. The article claims that the technique of goal crossing with their program is at least as expressive as pointing-and-clicking, and allows for extra features, such as single-stroke control, exploiting the strength of the stylus. For users of tablet PCs, a stylus or pen is generally used to help users with tasks that are normally done with a mouse on regular PCs, and CrossY does wonderfully in achieving fluidity in selection tasks. A pen is normally used in a flowing-stroke, as in drawing or writing, so to ask users to use a pen to act as a mouse, by using gestures like double-tapping as a replacement for double-clicking, is slightly awkward, even if the mouse-and-keyboard interface is widely-used. Providing the program to allow users to use goal-crossing as the selection method appears to be more intuitive for users. As shown by the researchers' own experiences and the guests in the open house event, the system appears to be well-received. One notable exception however is that the current system requires some explanation. The system is not intuitive enough to users, according to the authors, due to two reasons. The first is that the existing mouse point-and-click paradigm is already so well-established that the new system will be difficult to adopt. The second is that the CrossY system is currently not intuitive enough for users to explore, since most of the users in the open house event required some explanation in order to actually use the system. A more obvious visual display would allow users to determine affordances for the system, which researchers can use to develop a conceptual model, making the system more intuitive and easier to use. One suggestion may be to use a solid line for a not-yet selected option, and a visibly broken line for a selected option, much like a knife cutting through a vegetable. This suggestion also allows users to see where to move the stylus, unlike the current system which just has two difficult to see goal posts. Review of "Interacting with Paper on the Digital Desk" from The Communications of the ACM July 1993 by Pierre Wellner The article explains the implementation of a projected workstation onto a desk. It provides a short list of features that are provided by the system, and also presents a list of obstacles that a desktop projection needs to overcome in order to be useful. The approach appears to be an uncommon idea at the time the article was written. The technology at that point was mostly towards achieving virtual representations of physical materials (scanning documents into the computer, writing documents on the computer, etc.), but not the other way around. The "Digital Desk" strives to reverse the trend by providing an augmented physical space. This digital desk does not serve to replace the physical desk, but rather provide some additional functionality to a physical desk that one might find in a normal computer desktop. The display relies mostly on a single gesture that is very natural to a person, pointing. Pointing is undoubtedly natural to people because we are taught at an early age that a point serves to indicate others to direct their attention to an object in the pointed direction. With the camera, a computer can take the image and figure out if the user is pointing, and if so, where he or she is pointing to. Using this single gesture and a tap on the desk as confirmation, a user can do many things, like copying the number and inputting it into the projected calculator. The camera also can recognize a pen or stylus and its strokes on physical paper or on the display, and allows users to physically interact with documents similarly to a virtual display. Copying and pasting is the biggest feature that is difficult to do in real life, but quite easy for a machine; the projected display combined with image capture and gesture recognition allows a user to activate copying and pasting for a document, essentially creating a scanned copy that can continue to be projected. However, despite all of the useful features, a projected screen has many obstacles. To my surprise, this article lists many more issues with digital projection than it lists opportunities for expanding the digital-to-physical feature set. It seems the authors of the work believed that the issues that occur with the system will have a large impact on usability, so they believe research needs to be done on mitigating the problems first before expanding the feature set. The list includes problems like hand-obstruction, user handedness, difficulty in image capture, calibration, etc. It seems like more research is necessary to determine how problematic these issues will be to users, since it does not seem like the article makes a convincing case, since they seemed to have already suggested reasonable and simple workarounds.

phuongpham 23:00:59 9/21/2014

Intereacting with paper on the DigitalDesk: this paper introduces several interesting points about HCI using day-to-day behaviour by hands. With current technologies, reading the paper likes connecting the dots backward. The authors have proposed using many day-to-day behaviour by hands to maniplate with documents. We have seen these behaviour in today touch screen. However, many years ago, the authors have proposed to use these behaviour with cameras. I think the contribution of this paper is the Implementation issues. It shows valuable analysis for future research projects supporting hand maniplution with camera, e.g. SplitTouch project. This shows that even with the limit of current technology, research projects can always make contributions for future technologies. What I don't like in this paper is the morphology between the physical paper and the digital paper. We can using hand to edit a paper digitically. However, what have been changed can't reflect in time with the physical paper unless we re-print it. That is what touch screen technology can do by bringing the document completely digital but still bring the real feeling as a physical paper. ***CrossY: A crossing-based drawing application: this paper evaluates the feasibility of using crossing technique instead of point-n-click in a general application, i.e. sketching. The paper is another example of everything is best for something and worst for something else. Eventhough, the paper is not the first one presents crossing but I think the paper's contribution is using crossing technique in a full general application instead of a particular task as previous work. This shows the feasibility of replacing point-n-click with crossing technique. There are a few points about which I concern in the paper. First, the evaluated application is biased for pen devices rather than mouse, which raises a question when using crossing technique for all applications. Second, as the authors have said, the FlowMenu is not very fluence for user. I have never used it before but I can guess it will give discontinued feeling when I using it with a pen device. Third, the paper has not shown that the crossing technique can help pen devices over come point-n-click devices in touch screen devices. Last, I have a feeling that a fundamental with the current evluation is that crossing is a technique for pen devices but it was evaluated based on GUI built for point-n-click devices. Most of the evaluated tasks are just alternative to point-n-click actions on common GUI. I wonder if we can make a research on modeling new GUIs which are best for pen devices without related to traditional point-n-click GUIs. About both papers: We have read a few papers about HCI. The very first papers are seminal ideas which open a new research area, i.e. HCI. These papers briefly open research opportunities for later work. Recent papers go into depth research area in HCI. As the 2 papers we have read today, they proposed new way of interacting with computer using hands or pen devices. However, the reported experiments are somewhat unofficial where the authors reported their experiment result descriptively. I wonder if this is the way we will do in HCI area?

Wei Guo 11:48:58 9/22/2014

Reading Critique for Interacting With Paper on the DigitalDesk Wei Guo This paper introduce a technique called DigitalDesk, which can help interacting with paper documents on the base of computer. The DigitalDesk is a desk, on which we can put paper. It can project electronic images on the desk, interaction with pens and fingers, and also “read papers”. This are projector and camera on the above of the desk. There is also a digital tablet on the on the desk. User can put paper right on the tablet. Projector can project images, documents on the tablet. User can read number and part of the image of the paper by pointing it. Computer will stored the reading information and use it for calculation or for pasting and further edit. Two person can interact with each other on two separate DigitalDesk. We have to admit that Pierre Wellner is very talented to invent this around 13 years ago. I searched this DigitalDesk on the internet, and seems there are no such or similar product in the market. According to the paper, and video (http://www.youtube.com/watch?v=S8lCetZ_57g), DigitalDesk is a very useful and helpful machine. What prevents people from using it? Here are some of my opinions. 1. Cutting and pasting are based on image instead of text itself. 2. A projector, a camera, a tablet on top of desk is very space consuming, and expensive. Not portable. 3. When doing the sum, the user has to take out a card says “SUM ->” which is too complicated. Some commands are not attainable. 4. The technique DigitalDesk using makes it very slow. Even a simple “8+9” calculation question still takes more than 1 minute to perform.   Reading Critique for CrossY: A Crossing-Based Drawing Application Wei Guo When the input device is pen instead of mouse, traditional point-and-click does not work as effective as crossing. This paper introduces an app called CrossY, which implements the crossing interface. This paper also evaluates the CrossY application. The paper clearly introduces each corresponding functions of CrossY system to the traditional pointing-and-clicking system. The pictures give reader a very direct sense of how this CrossY system should be used. The authors evaluate the CrossY system in several ways: expressiveness, consistency, hardware and software considerations, and user feedback. CrossY paper won the “Best Paper Award” at the 17th Annual ACM symposium on UIST2014. I would like to analysis why this paper is good enough to get this award. First of all, this paper is very easy for readers to understand. The form of this paper is very formal. This writing of this paper is very clear. It doesn’t contain jargons to confuse the non-HCI reader. The logic is very reasonable. The demonstration of usage of CrossY system is very clear. Reader can easily understand the system usage with the help of the pictures aside. Second, this paper states a complete app instead of a simple idea and partial of implementation.

Qiao Zhang 12:39:37 9/22/2014

Interacting with paper on the DigitalDesk This paper proposes a digital desk, which can combine the two different worlds together: the electronic world of workstation, and the physical world of the desk. It allows users to work collaboratively by clipping a physical book page and share via the digital desk. An important point raised by the article is that interaction should try to leverage user's developed tactile skills. Virtual reality may not be able to replace physical tools. Hence, this paper introduces two desks: a paper pushing and a pixel pushing desk. In this way, the user can have a space that both electronic and paper documents are available. They try to enhance the paper reading with computers. The device can read paper documents placed on the desk, and project electronic images onto the desk/paper documents and can interact with pens or bare fingers. Example applications include a calculator, PaperPaint and the DoubleDigitalDesk. The calculator example does not appeal much to me. Although it indeed improves existent applications, because it is still imitating using a calculator. I have seen a very impressive app that utilizes handwriting recognition, and I believe this is more intuitive for human beings. The PaperPaint application seems good to me. It integrate the physical world with the virtual workspace, providing convenience for annotating and sharing. The last application, DoubleDigitalDesk, allows users share current workspace with each other. This reminds me of Google Doc to some extent. The digital desk reminds me of the Microsoft surface in the first stage. It comes with some problems because it uses projectionsI also found a similar commercial product: Epson X-Desk. It costs about 380 pounds, however, it does not come with the function of interacting with physical paper. I can imagine that just as mobile phones are once too expensive but now prevalent, personal projector-based interactive systems will become cheaper and more accessible for average people in the future. ============================================ CrossY: a crossing-based drawing application The paper proposes is a simple drawing application designed to demonstrate the feasibility of goal crossing as the basis of a user interface. The CrossY uses motion traces to select objects. Instead of clicking on buttons or menus, the users can cross the pointer over the objects. This approach can resolve the difficulty of using a stylus to pin-point a specific point. The paper uses the metrics that we just learnt from previous papers: expressiveness, efficiency, visual footprint and fluid composition of commands. However, if I am using direct manipulation such as using a stylus on a tablet, as a right-hand dominant user, I find it difficult to see contents under my hand. This will create extra footprint in addition to the stylus. This swiping approach can be seen everywhere nowadays. For example, I have used Kinect. They uses gestures for option selection as well, because it is nearly impossible to manipulate precisely on such a big screen using body motion. However, I do experience difficulty using it, because somehow it is very sensitive to movements. But it is fine with a stylus since wrist muscles are more flexible. However, one shortcoming for this approach is that crossing an item does not naturally indicate selection. It usually means deletion. Instead, I would suggest using circle motion to select an item, however this will add effort to the users, though. Another thing that I do not appreciate is the scrolling. The page up and page down gestures are not intuitive to me; as a user, I do not want to memorize unintuitive mappings between gestures and functions.

Nick Katsipoulakis 15:16:24 9/22/2014

Interacting with paper on the DigitalDesk :: In this paper, the design of a novel system is presented for combining desk physical tasks with electronics. DigitalDesk is based on the idea of extending capabilities of physical objects with electronic features. It works on the surface of a desk, where a screen is projected and input is parsed through cameras. Among many benefits provided to the user, DigitalDesk's main contribution is the ability for two or more people to collaborate on the same task. Furthermore, DigitalDesk provides a number of other features for ameliorating the difficulty of other desktop processes, like arithmetics. The feasibility of DigitalDesk evolved through solving problems for interaction, display of information and feedback. Even though feedback is the most important attribute for an interface, and usually systems strive for "direct manipulation", DigitalDesk follows the approach of tactile interaction. This is achieved through enhancing paper documents with additional electronic features and extending a system with new interactions (i.e. pen interaction, hand interaction etc). The display of information plays a significant role for the usability of DigitalDesk, since engineers had to work towards resolving problems that had to do with visual feedback to the users (i.e. shadows). Also, reading paper documents is another obstacle that engineers had to overcome during the development of DigitalDesk. Despite the fact that this system has not passed formal user evaluation, users had a positive impression about the system and the designers of DigitalDesk had the opportunity to reveal a number of caveats during testing (i.e. handedness, obscuring selections). ---------------------END OF FIRST CRITIQUE ------------------------------------------// CrossY: A Crossing-Based Drawing Application :: This paper presents the feasibility of an interface which uses the crossing gesture as the main unit of interaction. CrossY focuses on expressiveness (through measuring its performance with the WIMP interface), composition of different commands, efficiency and visual footprint. Its design consists of a menu for selecting tools and an interaction area. A detailed presentation of CrossY's features is presented followed by the design approaches for making a crossing-interface feasible. An interesting discussion on learned lessons of CrossY revealed seminal ideas about crossing interfaces. In terms of expressiveness, CrossY succeeds in covering a large part (if not all) of actions supported by traditional interfaces. Also, it is characterized by fluid composition of commands. Experimentation revealed that CrossY proved difficult in terms of usability for naive users. However, most of them quickly understood how CrossY works and found it intuitive and more usable is some cases.

Brandon Jennings 16:57:12 9/22/2014

Interacting with Paper on the DigitalDesk This paper is about combining features of both physical and virtual workstations. The DigitalDesk is a real desk, but enhanced with digital features like a computer display and cameras to give the user the best of both worlds. What makes this paper interesting and important is that the DigitalDesk is a great example of merging the physical world with the digital world. The cameras and projectors can send real world objects into the virtual world. Users can also share information from physical documents virtually with one another. Changes can be made to both the physical and electronic copy. What I would have liked to see was a formal experiment to test the efficiency of using such a desk. Test users can perform a list of tasks. One group can use traditional all-physical workstations and the other can use the DigitalDesk. Although in theory some of the features may seem more efficient, it could be the case that traditional methods perform better in practice. The level of tasks also needs to be tested. The DigitalDesk will be ideal in some settings and not in others. Maybe it is more efficient for simple and tedious tasks, but not for complex and detailed tasks. CrossY This paper introduces using pen-based interfaces as building blocks for interface systems. Other work as tried to justify using pen-based interfaces as a replacement for point-and-click systems, but the main problem is that pen-based doesn’t work well backed by point-and-click. With the increase of tablet use and interactive computers, traditional point-and-click methods are not efficient. A pen-based interface needs to allow for expressiveness and fluidity. The contribution of CrossY as a benchmark for crossing as a graphical user interface will become more important as technology moves towards that path. Crossing is a truer form of interaction as the user can have more freedom and more expressions to communicate to the device. Again, like the previous paper, I would have liked to see a formal experiment testing the efficiency of crossing against standard interfaces. I also appreciate the acknowledgement of the features mentioned in the future work, as I was wondering why some of them, like tactile feedback, were not already included in the system. Crossing has been shown to be a promising building block for an interface for devices such as tablets, as it is, in the worst case, as good as traditional point-and-click.

Bhavin Modi 19:06:12 9/22/2014

Reading Critique on CrossY: A Crossing-Based Drawing Application The paper can be summarized as trying to show how by using goal-crossing we can create a more fluid GUI. To demonstrate the author has created a simple drawing application called CrossY. In the previous paper Beyond Fitts’ Law, we learned about the models for trajectory based tasks and the steering law. Here we see a more practical approach. There is a clear need to work on this, because of the disadvantages in touch screens using point and select. Making the interaction more natural is the goal, we see the use in menus and toolbars. Experiments also show that this follows Fitts’ Law, but more experimental values would help in this article. Features like the find and replace using circular motions are indeed handy, the downside is that this not a complete approach. Other helpful features in pen based interactions could be better methods to copy paste, using a rectangle to copy the area you want. The problem with such an approach is the interfering objects in most cases. Also we will have to revert back point and select, unless we want some unwanted lines in the drawing. The users have to remember many gestures which is a problem and frustrate the user with errors due to an incomplete gesture or a wrong motion. The question is why draw gestures when you can do it normally, as such we can say that the goal based approach is not viable over the entire design space. It is true we want fluidity, but we cannot have continuity and at the moment for the current interface it is available only in chunks. The spatial layout as mentioned is the key factor when we plan to use goal crossing. --------------------------------------------------------------------------------------------------------- Reading Critique On Interacting With Paper on the Digital Desk The Digital Desk has been discussed in this article. The working, the need, the thought process, current flaws and future enhancements. The main idea as mentioned by the author is to enhance the physical world by providing it better access to the electronic world. The article discusses the digital desk at length. To begin with the advantages of the desk we have to understand that people are more comfortable using physical objects, analogy to a desk in the office, hence the name. To be honest, I too prefer reading from paper and directly manipulate objects, the mindless doodling is what I miss. All this makes the physical more comfortable and natural. So instead of having a workstation on a desk taking up space and switching back and forth between them, the author suggests the Digital Desk. The concept that most intrigued me was the idea of collaboration. To be able to work with people worlds apart in real-time. I believe there is a lot of scope for research and new innovation in this area. The thought and the features at that time are truly amazing. I say add a printer to the desk, and we can scan using today technology. We will have something that will truly change how we work. The paper provides new direction of thought than the previous one. They focused more on the computers and how to improve the use interface for better performance. Here we are thinking how to create a collaboration between the physical and computer, and how to enhance the physical so as to develop better smoother user interaction. Many of the problems mentioned have been overcome today. A good example of the modern Digital Desk is Microsoft Surface Table. The new technology removes the need for projectors, the shadows are gone, it is Multi touch detection and touch screen. All this is possible using pixel sense technology, each pixel is actually a camera. We have come a long way from what is described in the paper.

Qihang Chen 19:14:38 9/22/2014

Interacting with Paper on the Digital Desk ----------------------------------------------------------- In this paper, the authors introduce the "Digital Desk," a real physical desk on which users can interact with real world objects such as paper, pens and pencils, but that is also enhanced with electronic technology. All the collaborative features mentioned in this article exist such as collaborative editing (Google Docs), audio/video conferencing (Skype), and desktop sharing (Skype, Google Hangouts). The other features that involve bringing digital properties to physical documents also exist in one form or another. With modern handwriting recognition technology, a stylus and tablets can be used just like a pen and paper with the added benefit that the computer can spell-check, clean up ugly handwriting, and allow dynamic changes. Additionally, images of physical documents can be converted to a format that allows the user to edit the document digitally and then send it to people, and other machines. It’s hard to say if modern tablets are still reliant on metaphors to real life or if they enhance the abilities of physical documents. It could be that modern developments are pushing from both sides, making machines more like real-life and real-life more like machines. This article presented many concepts, unfortunately it was written in the wrong decade. --------------------------------------------------------------------------------- The paper on “CrossY: A Crossing-Based Drawing Application” is about a simple drawing application developed to test crossing as a valid form of human computer interaction. The authors discovered that that crossing can be just as effective as point-and-click interaction, and that it also allows for more flexibility in interaction design. It turns out that I still prefer the print version, and it is for the reasons that the authors proposed; I like the tactile feedback and the easy manipulation of the paper copy. The digital desk, therefore, might have been a way for me to save paper, and until I can get something like it, I do not think I will move to the all-digital paper reading. As for their projection problems, if this paper was written today, I think the authors may have focused on more touch based technologies. As soon as I read about their shadow and location issues, I thought of a giant iPad as a table. It senses touches, removes the problem of shadows, and adds color to the desktop, as well as an all-digital display. If the table could read documents (i.e. scan and display), the need for overhead cameras would be reduced, which would minimize the footprint of the device as well.

Yanbing Xue 20:51:50 9/22/2014

The first paper demonstrates a very interesting system. It combines the two interaction styles, the electronic world and the physical world. Each world has advantages and constraints that lead us to choose one or the other for particular tasks. Interaction with documents in the electronic world and the physical world both has its advantages and disadvantages. How to take advantage of these two kinds of interactions becomes a problem. There are two directions. Researchers prefer either to transfer the digital equipment to the metaphor of real world objects that people are used to, or to do it in the opposite way. The DigitalDesk works by mounting a video camera above a desk. The output from the camera is fed through a system that can detect where the user is pointing, while also using optical character recognition (OCR) to read documents that are placed on the desk. Since most humans are fluid when using a pen, moving from a pen type of interaction to a point and click type of interaction on the same device can get sloppy and slow users down. One of the broader goals of this paper was to introduce a way to replace this point and click interaction on tablet devices with a penmanship style of crossing. They stated that they lose no expressiveness over the traditional interactions with this technique, and are able to keep the interaction consistent. With PaperPaint, the cut and paste metaphor can be applied to real world objects speeding up creativity. With two Digital Desks, people can collaborate simultaneously on a document in ways not possible with traditional paper. Many more uses are possible, and only limited by imagination. However, it has many disadvantages. On a workstation, we can obscure something with the pointer as we select it and the system still knows what is underneath. When pointing at paper with the DigitalDesk, the system must be able to see the paper and this means that fingers and other pointing devices must be out of the way. Besides that, when people work on the paper, it is flexible to make a move or to rotate paper for certain purposes. Using DigitalDesk, it obviously cannot meet these requirements. Except for these possible disadvantages, the invention itself is encouraging. It at least enables to use paper documents to perform useful tasks that are more awkward to do in other ways: tasks such as copying a long number into a calculator, translating a foreign word, replicating part of a sketch, or remote shared editing of paper documents. The second paper is mainly about a simple drawing application designed to demonstrate the feasibility of goal crossing as the basis of a user interface. Crossing uses steering to select paramaters. Instead of clicking on a radio button, you cross the pointer over the desired parameter. This application implements a number of different controls using this technique. While certain menu designs are based off of FlowMenu, many of the other elements in this system are new. While Crossing was previous identified as a potential substitute for the classic point-and-click interaction, this work is the first to report on practical aspects of implementing an interface based on goal crossing as the fundamental building block. In recent years, the portable, pen-based computers has been used in many cases. Many WIMP interactions that were originally developed for mouse are difficult to perform with a pen on a tablet computer. So the author explored crossing as a fundamental building block of graphical interface interations. The design goals includes expressiveness and fluid composition of commands. The exploration of this technique has been proved to be successful. And very interestingly, not only with the pen-based system the authors supposed, it is proved to be very successful with today’s smart phones. Many of the other difficulties with the DigitalDesk can be attributed to the technology of the time, such as the limitation of the camera’s resolution, calibration and touch detection. But the DigitalDesk does have many advantages not found in touch devices of today due to the projection of its display. Overall I found this paper to be interesting in creating a new GUI toolkit for user input. I am somewhat surprised that this idea of crossing has not been adopted, if not only in drawing applications.

Longhao Li 22:07:27 9/22/2014

Critique for Digital Desk This paper talked about a new way of interaction between human and computer, which is trading real desk as computer interface. Human behavior on real desk will be used as operation of computer. Basically, a digital desk is a device that projecting a graphical interface on a desk, and people can use actions on real papers to operate computer. People can draw pictures on paper and transfer the image to computer by just placing them on the projected interface. The camera above the desk will take care about the input of images. The interface also supports multi-users’ online corruption. People can draw pictures and manipulate documents together from two digital desk devices. This is a great feature that gave people the way to use computer without learn how to use it. Also this interface didn’t sacrifice a lot of feature of mouse keyboard operating interface. It will be very useful for some particular users that manipulating a lot of documents and pictures on computers and papers. Even though that the interface bring us a great approach for human computer interaction. It seems still has some problems. Everything is good for something and worse for something else. For people who mostly working on drawing, I think this interface are the most suitable one. It is easy to use and they can direct use the on paper drawing skills they have learned on computer. But for people who mostly work on typing. The interface will not be a good approach. Since people have learned how to use keyboard frequently to typing things (faster than hand writing), using this interface will slow them down. Even though if they can have virtual keyboard for typing, still the feeling of typing on desk is different with typing on real keyboard since there are no feedback in the desk interface. So still this interface cannot replace our old fashion way of computer operation. But in some field it can be the best choice. Critique for CrossY This paper mainly introduced a new graphical interface which use pen instead of mouse, and use cross gesture instead of strokes only. Based on using of different cross gestures, users can achieve most of the operations that we can do on point-and-click interface. This paper gave us a new understanding about how to design non-point-and-click interface. I think it is an important paper. Firstly, it followed people’s natural willing to use pen on papers: doing fluid drawing instead of doing strokes. People will not take a long time to learn the basic gestures to operate the interface. By fluent using the gestures, it will be easy to speed up the operations. But it still has problems. Since this system mainly use cross action, some gestures are designed not so intuitive, like page up and page down. People will not do that in nature way so that users need to learn and memorize the gestures. It will lead to inconvenient for users. But just like the author said that this interface is aim to test how cross gesture can improve the operation of tablets, and as result, I think it works great. In nowadays’ tablet interface, cross-gesture turn out to be used a lot for computer’s operation. It definitely helps users a lot on the operation. For example, people use swipe left and right to change to next or previous picture in their album, and in some operating systems, swipe left or right start from edge of the screen can act as forward and backward between webpages. All of them make the operation easier. If we don’t have these features, the operation on tablet will be hard since users need to keep stroking on the little buttons on screen, which needs careful operations. I believe with the development of mobile device operating system, cross interfaces will become more powerful, and the operation will become faster and easier.

yubo Feng 23:05:02 9/22/2014

There is 5 articles in today's reading section and 4 of them are talking about hierarchical marking menus, all of them proposed that inflection-free simple marks is much better than single zig-zag mode. In paper "Simple vs. Compound Mark Hierarchical Marking Menus" the authors try to compare two different menu select mode, which shows that former is more better, or could not be worse than the latter. In the paper "The Limits Of Expert Performance Using Hierarchic Marking Menus" proposes the concept of pie menu, which is the source f hierarchical marking menus. In the paper "More than dotting the i’s — Foundations for crossing-based interfaces", the author try to discuss two methods, crossing boundaries and selecting the performance which is better by redefining Fitt's law. All of the papers discuss the input performance could be different from method to method, and turns out that, the most future way is the way that consuming less action, I think it is most like that the short-path effects in interface performance.

changsheng liu 23:10:56 9/22/2014

<Interacting with paper on the DigitalDesk> introduces a physical desk which is enhanced to provide some property of electronic workstation. The DegitalDesk has a video cameras to record what the user is doing and uses image-processing techniques to recognize it. So I think the system has a high requirement of the gesture recognition algorithm. It also has a projector to display the images onto the desk. User can interact with the system using pens or bare fingers. The proposed GigitalDesk is a good way to interact with paper in some cases, such as copying a long number into a calculator, translating a foreign word, replicating part of a sketch or remote shared editing of paper documents. I think it's very challenging to implement for the following reasons:(1) The cost is very expensive. It has a projector and camera(maybe more than one). (2) The accuracy of gesture recognition. <CrossY: a crossing-based drawing application> explores crossing as an essential building block of graphical interface interaction. The authors developed CrossY, a simple sketching application in which all interactions relied on crossing. In the paper, some unique challenges are provided. The reason for crossing operation is that in graphic drawing application on tablet, pen use encourages a fluid and continuous style of interactions, whereas point-and-click interfaces insist on segmenting interactions into a series of pointing steps. The authors did some experiments to demonstrate the four aspects of crossing: expressiveness, fluid composition of commands,efficiency and visual footprint. It turns out that CrossY is at least as expressive as the standard point-and-click interface. It's possible to compose commands in one single stroke. Regarding the efficiency, it shows that for novice users, the space requirement of a crossing interface will be similar to the equivalent point-and-click interface. Finally, based on the experience, a slightly larger footprint may be acceptable as the expected speed benefits from command composition are substantial. In fact, the paper cannot persuade me that crossing is more efficient. Even it's can decrease the time for selecting a color, item or size, it cannot improve the efficiency dramatically. This is because in drawing, people would use most of the time to think about how to design the picture, not selecting.

Xiaoyu Ge 23:36:28 9/22/2014

Interacting with paper on the digital desk This paper introduced a human-computer interaction different from the normal design making the digital interface analogous to physical world. The design is to make the physical word gain more electronic functionalities. This idea is pretty brilliant, with more computational abilities the physical items not only function as it used to be, but can also having other functionalities which can be more easily accept by users and also make the item more useful then ever. This kind of HCI design method is widely used now a day in industry. Google glass for example, is developed to have functionalities as a mobile device which have installed a android system in it, which make the glass be able to make phone calls, take pictures and exedra. However, just as the paper illustrated the form of glass cause limitations to the performance and user experience problems of the extended new functionalities. Moreover, watch has also been redesigned and developed to having more computational abilities. Apple and Google both have published their smart watch, which not only have traditional function of showing time but also be able to be used as a small computational device. CrossY: A Crossing-Based Drawing Application This paper present an exploration developed a Crossing based graphic user interface, which is as expensive as the traditional point-and-click interface but provide more flexibility and encourage the fluid compositions of commands. Even though the crossing based graphic user interface is beneficial for pen-based interactions on Tablet-PC, there are still software and hardware limitations such as unable to track pen outside the screen and user start crossing gesture before landing the pen. And the paper manage to solve these problems since there are new hardware model solved this problem and author queued pen positions of flying pen gestures which also solved the software problem. The cross-based interface design method seems pretty good for tablet users who heavily used pen but I have never seen other applications use the same kind of interface design method. It seem that the crossing-based design use different gesture to achieve the same function as the traditional point-and-click interface which might confuse users who have be used the point-and-click for years. There are too many new gestures, which is not easy for user to get familiar with. Even though there are a lot of benefits, users will not willing to learn new gestures just to use only one application when plenty similar applications designed with tradition point-and-click interface are there in the market. Moreover, a successful human interface design should reduce the effort of user to perform the desired function. This approach failed to achieve this basic goal.

Mengsi Lou 0:09:43 9/23/2014

Interacting with paper on the DigitalDesk -----------This paper introduces a new design of Digital Desk that is a balance between “paper pushing” and “pixel pushing”. The digital desk is a real desk that people can use it as ordinary desk, and it also contains display on the desk and can capture users’ action which makes the interaction come true. -----------Following I would like to focus on the implement of this digital desk. The first thing is the interaction on the desk. Beyond the traditional pattern of “direct manipulation”, it offers interacting with electronic objects. One way is video-based finger tracking. But in some situation finger is too thick. So a more efficient way is to look for motion, that is to pick out moving objects from the background. Another problem is hard to determining users’ taps on the desk because of the overhead camera. One way to solve is by listen. The second point of design is the output. A potential problem is the shadows. But it turns out to be fine when in real use that people may not aware of the project overhead because they often get used to the lamps that also brings shadows. And brightness of the room is another issue with projection. The third step is to look at the user experience. Feedback is always one of the most important things. ///////////////////////////////////////////// A Crossing-Based Drawing Application ---------This paper tells about a study of crossing using in graphical user interface. The standard WIMP-interface and the traditional point-and-click interfaces are not well adapted to direct pen interaction. Thus, comes to the crossing pattern that is suitable for pen-based and mouse-based interactions. The goal can be divided into expressiveness, fluid composition of commands, efficiency and visual footprint. The author develops a application called CrossY that is a sketching program offering tools including a pen, an eraser and a highlighter. CrossY also offers a simple search-and-replace feature that will make the users change attributes easily. A typical move in selecting pen attributes is to open the pen attribute dialog box by crossing the pen tool button and extending the stroke towards the left. -----------This paper reminds me of the convenient finger gesture control on iPad and also on the touch tablet on Mac computer. They also break the traditional way of control and interaction and through some multiple fingers gesture make the interaction more efficient. At first users may not get used to this pattern because of early use of Microsoft software and the use of mouse. But later it turns out to be more efficient and even do not want to use mouse anymore when not operate accurate moves.

zhong zhuang 1:36:54 9/23/2014

This paper discusses an implementation of digital desk, in traditional technology, the paper world and digital world is usually completely separated, there are properties of each world that are so powerful that we can’t afford to totally shift to one world. For the paper world, paper is tactile, can afford direct manipulation with pen, marker and eraser etc, paper is super cheap, has high resolution. For the electronic world, documents are very easy to edit, copy and share, so if we can combining these advantages together, we can significantly improve our document processing system. In this paper, the idea of direct manipulation is highly extended; it adds electronic properties directly to paper. By doing this, we go beyond the size of screen, we abandon intermediate tools to operate documents. As proposed by this paper, by adding a calculator on the desk, user no longer need to key in numbers from paper, but directly drag and drop them. By using PaperPaint, user can easily duplicate their sketches just like they do on computer screen. By DoubleDesk, people from two countries can operate on the same paper document. These valuable features improve our user experience significantly. I believe this kind of direct manipulation interface will become the next generation of human computer interaction tool.

SenhuaChang 1:44:51 9/23/2014

Interacting With Paper on The Digital desk This article is mainly about a new implemented system called Digital desk which can make the desk more like a workstation, I like the sentence which is the first sentence in the conclusion of this article-”Instead of making the electronic workstation more like the physical desk, the digital desk does the opposite: it makes the desk more like the workstation”. I like this sentence because sometimes, a good idea comes from breaking the stereotype, which means think the same thing in another side. At quick scan, I think this article is like an introduction paper of this new tech. The article explain how it works and how to solve the problem during the process. There are a lot of photos in this paper which give reader an intuitive feeling and knowing of the tech, and it support author’s explanation about looking down to the desk and the glass problem. In end of the critique, I have to say Digitaldesk is an amazing project, which could be the considered as the basis of many work we see nowadays that aim to work together with both physical object and virtual object seamlessly. However there still exists some problem, which should be taken in to consideration after doing the whole user experience and do a better products. 09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)09:47, 23 September 2014 (EDT)~~ CrossY: A Crossing-Based Drawing Application This paper use a drawing application to demonstrate to the reader the advantage of designing user interface based on goal crossing. This article argues that crossing is as expressive as the traditional point-and-click interface and provides designer with more flexibility because it contain shape and direction of the strokes. This article also shows that crossing is not only good for pen-drawing but also good for pointing which like the traditional interacting way. In my opinion, I think crossing is good at some place such like drawing task, but in any other place, more experiment and more useful toolkit of widgets should be provided such like what author said in future work. Only based on this paper, I am not convinced that it can do better than point-and-click, after all, most of us are familiar with that.

zhong zhuang 2:08:09 9/23/2014

This paper proposed a new basis of a graphical user interface, in recent years, tablet and smartphone is more and more dominant than personal computer, people will spend more of their time on tablet devices than PC. In this situation, traditional point-click paradigm shows tis limitation, because we use pen and finger instead of a mouse to interact with the device, compared to mouse, pen and finger is a much “coarse” tool, it is difficult to perform certain tasks using finger and pen, for example, double click and select a line of word, etc. This paper introduces a more fluid and continues way of interacting – crossing. Instead of aiming the pointer to the target and click it, user will just cross the target to select them. This is a more natural way of interacting with computer using pen and finger. But this also completely change the way of HCI, so lots of concerns have to be taken into consideration. As discussed by the paper, first we need to consider can this new language express at least the same set of features than the current one? Is this new language efficient to implement and to use? Is this new language intuitive to novice users? Does this new language require very large screen size to implement? These are all questions that need to be studied very carefully. The paper examines the new language by implementing a prototype, CrossY. A drawing tool based only on crossing. It illustrates varies components of the application and explains how user can interact with them using crossing. From my point of view, I think this method is more efficient than point-click, because user can make multiple actions in one stroke, and it is easier to use by using pen and finger. But I think it is not intuitive at all. New users will take huge amount of time to adapt to this new language of interaction.

yeq1 6:23:30 9/23/2014

Yechen Qiao 9/23/2014 Interacting with paper on the DigitalDesk In this paper, the author introduced DigitalDesk, a novel device that attempts to further blur the lines of physical and virtual desktops. DigitalDesk uses a projector to display the virtual content so it can be combined with physical objects, two cameras that are used to capture the user’s input as well as scanning the documents, and networking support so people can work on the same digital content together at separate locations. In many ways, DigitalDesk had been ahead of its time. The main idea of bringing physical objects properties of physical world was novel and definitely useful. Due to the technological limitations, the author had to implement this device with technologies that are less than idea. Without affordable high resolution flat panel displays, the authors had to use overhead projector which causes some kind of shadows (though users seems to not be too bothered with them). Without affordable high quality digital cameras, they have to use two cameras to scan the documents. They also requires users to keep the documents still when scanning. While this prototype does not allow direct local multi-touch, it allows other users to touch on the “virtual desk” remotely. However, the study had showed that despite all these limitations, the new interaction already appeared somewhat appealing to the users. However, I do wonder how much this prototype would cost at that time… There are potentially other interactions the author did not explore with this device. For example, having two cameras would allow the device to see the images “3D”. Is it possible to support holoprojections and allow the users to edit 3D models as if they are sculpting? What about allowing detection of pressure sensitivity on the desk, so we know how heavy a stroke is for the paint program they demonstrated? Today, we have similar devices appear in the market such as Microsoft PixelSense. PixelSense shares many similarities with this device. For example, it is a table with five cameras that can sense the user’s input. In theory, the users can use Surface 2.0 SDK to develop applications that scans the document as they did with DigitalDesk. PixelSense is slightly more advanced, as it uses high resolution (1920x1080) 40’’ flat panel display, the IR cameras allow the system to detect different types of real world objects. For example, the paint application could detect whether the user is holding a pen or a brush, and adjust the input accordingly. The new technology also allows multiple people to operate at the same desk. However, the device is expansive even today. Even with improved technology, the second generation PixelSense is still selling for around $8000 minimum. (https://www.youtube.com/watch?v=kmOku92MlQc Also, it looks like someone is selling one in eBay for $5000) I speculate most of the cost is due to the price of the huge flat panel. I think by the time we can purchase affordable 4K displays, devices like this would start to become affordable as well. CrossY: a crossing-based drawing application In this paper, the authors have described CrossY, a prototype drawing application with crossing interface designed to facilitate better pen inputs of commands while still provide adequate support for mouse input, and allows for command compositions. The authors evaluated the expressiveness, and argued why composition of commands is useful. While the authors did not formally prove expressiveness, I can see the states (idle, pressed, release) of point and click to be mapped to (idle, crossed, release) and maps between actions of each state (click while idle->cross while idle, drag while pressed->drag while pressed, release while pressed->release while pressed). However, this assumes we have infinite dragging area. If, for example, we are required to cross from right to left, and I want to drag a window’s title bar (assuming “drag” bar spans the entire title bar, with no minimize, maximize, or close buttons on the right, like Windows 3.1) to the left with a pen, I would be unable to hide the same amount of window content with CrossY as I would with point and click (I’ll buy the difference is small (there are around 1-10 pixels you will not be able to get since that’s being used for crossing recognition) and potentially not noticeable in small screens), and my input parameters with mouse clicks would be reduced by a bounding box of margins around 1-10 pixels for crossing recognition as well. Yes, these are minor differences, but I don’t think the authors can say CrossY is at least expressive. The authors had also explains why fluid composition of commands are useful. I personally would buy this, but I would also comment that the author did not take into account of “human error”. Suppose I’m in the middle of executing a composition of commands with a pen. The command takes 10 steps, and I made a mistake on the 9th step. Is it possible for me to recover from this error without having to cancel the entire action and re-execute these 9 steps again? What is the expected performance when assuming the user has (1-e) chance of successful execution of each step? Is there a tradeoff so for each number of compositions n, there’s E so that whenever e > E, we would have composition’s performance worse than that of multi-level? My last command is that I think the authors might want to make some minor edits of some actions of CrossY. I did a trial of the interface with pen using the PDF reader. I found the “find-replace” action particularly cumbersome to execute: 1. The “Find” cross is way too small and the distance to the last menu is too big. The angle of entry is too big for some brush combinations. It is difficult to find the right vertical stroke at such a long distance ahead of time. Either make the distance shorter or make the cross larger. 2. Once we get to replace, the problem becomes much worse. I find my hand constantly obstructs the next menu item I’m trying to select. I constantly overshoot the next menu item because as soon as I saw the next menu item, my stroke is already way past it. I would encourage the authors to do a comprehensive user study, and learn what other problems this interface can have and try to make it better.

Christopher Thomas 6:26:17 9/23/2014

FIRST PAPER ***** 2-3 Sentence Summary of Interacting With Paper on the DigitalDesk: The authors describe a digital desk, designed to make a desk more like a computer. Through the use of projectors and cameras, users of the desk are able to interact with real paper, with the added benefits of features such as copy/paste and the ability to “write” on real paper and to move the selection. The desk even supports a “teleconference” ability where multiple “desktops” can be displayed and users in a meeting can see everyone’s desk at once on their actual desk surface (not virtually). When reading about the DigitalDesk, I was very much reminded of a paper I read recently about TelePresence, which is a currently active research area investigating how people can be virtually present with one another. Obviously modern technology is more advanced than this system, but it is amazing how much the DigitalDesk was able to accomplish all the way back in 1993 – an era of dial-up internet and frustrating technologies. The DigitalDesk was truly forward looking in that regard – particularly in the ability to have group meetings with multiple digital desks. That was the feature that impressed me the most in the whole paper. We all take for granted nowadays that meetings can be held remotely and in groups over the Internet. However, back in 1993, this technology was just starting to develop. The closest you got to remote meetings were conference calls back then. I can see how the DigitalDesk is a great example of a direct manipulation interface. For one thing, the DigitalDesk actually functions in the physical world – with real paper. It allows people to bring some of the conveniences of working digitally to the medium which they are most comfortable with. People can write directly on paper with their favorite pens and pencils, copy what they’ve drawn, etc. It also serves to bridge the gulf of execution for some tasks as well. The article mentions how, for example, a user could simply place a piece of paper on the desk, point to a number, and have that number placed into a calculator without having to transcribe it. One criticism I would have of this is that the authors sell this point by stating that not only is it easier to just point, but also it will reduce errors because many users are likely to make errors when transcribing numbers from one medium to another. I’m not sure if I completely agree with their statement of this, especially not in 1993. How accurate could the detection of the numbers on the printed page be (character recognition)? Is it really more accurate than the errors people make in transcribing the numbers into the computer? So, while I definitely think it is a good idea and probably saves time, I would like to have heard more about the accuracy of their ability to place numbers directly into a calculator by tapping the finger next to it. All in all, I think the DigitalDesk was an innovative exploration and an example of a physical direct manipulation interface, which allows users to actually use their real desk, with real paper in an extended way, by turning on a projector. Obviously, further work was certainly necessary (inability to move paper at all, etc.) but I believe this kind of technology is still being developed today – just in a different form. Consider the Microsoft Surface technology developed several years ago. Microsoft’s Surface in the Home product allowed users to directly cut food on the surface of a touch table and showed them step by step how to prepare food – even telling them where to lay the food on the touch table. The product obviously was a huge step forward from the projector interface – supporting touch many new features without using cameras to detect fingers and sounds of tapping, but if one really thinks about it, the DigitalDesk was a sort of predecessor to that technology. It is interesting to sit from our position today and see the roots of these technologies.   ***** SECOND PAPER ***** 2-3 Sentence Summary of CrossY: A Crossing Based Drawing Application: The author presents a paint application whose GUI is manipulated entirely through a “crossing” gesture (crossing over with a pen and dragging). The author demonstrate that crossing is at least as expressive as traditional point and click interfaces and has the benefit of being more intuitive and more user friendly. I have not ever owned a phone or tablet where the interface was primarily through a touchpen based interface, so I don’t have a lot of experience with pen gestures. All of my touch devices allowed direct input with the finger. However, I thought the idea of crossing was interesting here and intuitive for some of the tasks the author demonstrated it for. Still, I think it is important to note that while crossing may be a great technique, it may not be the best technique in every situation. For example, when the author showed the example of exploring a directory hierarchy with the crossing technique, it simply looked time consuming and clumsy (and unnatural), particularly if there were many folders / files with different names and nested. This would require numerous gestures to be made over and over again over the file names. This could easily cause mistakes if the user is not careful and inefficiencies. I think it may simply be easier to show a traditional user dialogue in this situation and allow the user to simply point to their intended folder to open it. There is a famous saying that, “When the only tool you have is a hammer, everything looks like a nail.” In the case of this paper, I see that the author has demonstrated that the crossing gesture is rich and capable, but I think the author would have been better off demonstrating how the crossing gesture could work alongside other gestures better suited to the task at hand. I found the argument given in favor of the directory hierarchy exploration actually quite weak. The fact that no formal user studies were conducted raised my eyebrows as well and makes the author’s claims seem somewhat specious. In addition to the above, I also think that the author didn’t motivate the concept of crossing well in many of the contexts in which he applied it. In the cases where a user was simply picking “Find or Replace” or check boxes, wouldn’t it just be easier than have the user actually have to cross over their selection? I would argue the more natural and less invasive gesture would simply to have the user point at it with their pen. In fact, the author even mentioned than many of the users in the informal tests where users tried the system found many features non-intuitive or needed help discovering the more advanced features (such as the directory navigation). On the other hand, I think the concept of utilizing the crossing gesture in the right situations is a great idea. For instance, positioning the brush palette such that the color and width are able to be selected in a single stroke is a great idea and saves time for the user. At the same time, this may be inconvenient (if you look at figure 5) because if one wants to select a green color and a very thick brush width, it may actually be more difficult because it involves navigating the space between the two bars to find the preferred thickness, without overlapping the thickness and selecting the wrong one. So, while I think the idea is good in principle, only experimentation would reveal how it would function in practice, particularly the directory hierarchy navigation part.

Xiyao Yin 6:54:11 9/23/2014

‘Interacting with paper on the Digitaldesk’ provides us a new combination called DigitalDesk to make electronic workstations analogous to the physical desk, users can take advantage of their knowledge of the physical world. DigitalDesk is a real physical desk which is enhanced to provide some characteristics of an electronic workstation. It projects elecronic images down onto the desk and onto paper documents, responds to interaction with pens or bare fingers and can read paper documents placed on the desk. DigitalDesk is new in calculators by providing another means of entering numbers and makes it possible to select and paste paper documents in the same way like electronic documents in paperpaint. What’s more, DoubleDigitalDesk makes it possible to share real paper documents. Although there are still some challenges, DigitalDesk makes the desk more like the workstation, and it supports computer-based interaction with paper documents. However, what impressed me most is in the last part of this paper. Do people work primarily in the computer but with access to physical world functionality or do they work primarily in the physical world but with access to computer functionality? It is a good point to be discussed with. It seems that many research on computer are just in a wrong direction. People should make computer work in our real world, not make ourselves work in the computer’s world. ‘CrossY: A Crossing-Based Drawing Application ’ describes a substitute for the class point-and-click interaction. CrossY is a simple sketching program offering several tools. It is new in this area because it is the first to report on the practical aspects of implementing an interface based on goal crossing as the fundamental building block. Although the point-and-click interface has been very successful for desktop computers, it is not well adapted to pen-based interactions. In fact, I have the same experience. I am now using Wacom Tablet. It is easy to paint on the computer. However, it is hard for me to click on one item, in fact, I never succeed to finish a double click by using the pen of the Tablet. It seems significant to have a new technology in replace of the click process. After I read this paper, I find CrossY may be a good solution to this point. This paper compares using crossy and the standard point-and-click interface in different aspects. It shows that crossing paradigm is at least as expressive as the latter and provides the same level of functionality. I strongly believe this will ultimately lead to more efficient and natural interfaces for pen computing.

Jose Michael Joseph 7:51:01 9/23/2014

Interacting with paper on the Digital Desk This paper is about reversing the current trend where the workstations are built modelled after desk, in this approach the desk is modelled after the workstation. The idea behind this is that since we have innate abilities such as holding and writing with a pen, a workstation could be built that uses your innate and intuitive abilities. Thus what this system tries to do is simply put as build a workstation which uses our already learned abilities as input mediums. One of the earliest examples that the author provides is that of the machine called Xerox PaperWorks. What this does is that it allows ordinary papers forms to control a PC through a fax machine. The drawback of this approach is that a fax machine is much slower as compared to a PC and thus drags down its performance. Next the author discusses DigitalDesk and states its purpose as to support rapid and direct computer based interaction with selected regions of paper documents. While the purpose might be served with the implementation it must be noted that the DigitalDesk is a huge system. To achieve the goals of this system we require a lot of components such as cameras, projectors etc. In this aspect I think that the DigitalDesk is not efficient as it does not provide features that are worth the cost of the equipment. Also the DigitalDesk in my opinion is not rapid enough to handle all paper basked tasks. For example, suppose there is a book that we want to copy a part of. A DigitalDesk will take considerable time to scan each page and load it into the memory where we can use it effectively. The DoubleDigitalDesk states its primary advantage as that of two people living in different continents to simultaneously see and write each other’s papers. While the idea behind this has been unique, the implementation in this case is too costly to provide any significant benefits. It would be much easier for the two users to use a product like GoogleDocs to communicate and work on a single project at the same time. Thus the efficiency of a product like DoubleDigitalDesk is seriously questioned. Also the DigitalDesk has far too many variables that can affect its performance. Some of them are shadows and brightness. The shadows of the user’s hand could come in between the system and it’s ability to process the shown information. In a similar way brightness too can affect the performance of the system. I believe that in our current age when everything is moving from paper to digital, a system that tries to bring things back into paper is futile. It would be much more cost effective and efficient to have a software implementation which takes the e-copy of anything such as a bill, book etc and works on it to produce the desired effect. We are in an age where we have touch screens that can be written on like it’s a notepad, we no more feel the necessity to have paper based systems and thus a system like DigitalDesk is not required.

Jose Michael Joseph 7:51:36 9/23/2014

CrossY : A crossing based Drawing Application This paper discusses a drawing application called CrossY developed to show the feasibility of using a pen based input medium. It goes into lengths about the various aspects that using a pen as an input device brings into the picture. Their work starts off by saying that it demonstrates the feasibility of crossing as an interaction paradigm and also provides initial feedback on the unique challenges of developing such a system. Thus this paper only goes on to state the challenges faced during the implementation of the system while offering not too many comparisons with the current systems making it hard to gauge the effectiveness of the said system. It states that pen use encourages a fluid, continuous style of interactions. But while this is true it does not mean that a fluid continuous style is better than a discrete and abrupt style. There could be many situations where a discrete abrupt style can perform a task much quicker than a fluid continuous style. Thus again due to the lack of comparison in this manner we are left in doubt about the feasibility of this system. The paper does states that this system can be quite cumbersome and far less efficient for directories containing large number of items. It can be correctly inferred from the current trend that directories and databases are only ever increasing. Thus a system that does not perform well with a scalable directory and database structure would be inefficient in the current age. It must be noted that the paper clearly says that even though the idea was quite simple, its implementation had proved to be very difficult. This raises questions about the feasibility of such a system when we raise the features in it from basic to those that are currently found in its modern counterparts. Even if such a system was feasible, by what we can infer from the paper it can be stated that it would take quite some time to develop this system, which would again make it inefficient as by that time the technology and current standards would have changed rapidly. In all the comparisons that the paper makes between point and click and pen based input, it seems to only consider novice and naïve users. It is quite critical and necessary to consider proficient users and compare the statistics based on their use because eventually every naïve user will become a proficient user. Yet this paper has no mention of any sort of comparisons with respect to users who know how to use the system well. In fact it states that even naïve users needed some explanation to figure out how to use the system thus proving that it is surely not intuitive. Another comparison that was vital was the one with regards to menus. The paper had discussed the idea of navigating menus using a pen as an input device but failed to provide any comparison on the time taken to traverse a menu using a mouse and time taken to do the same using a pen. This lack of comparison has left a major question hanging with respect to the feasibility of this system. In conclusion I believe that if this system does prove to be even close to as efficient as the mouse, one of the places where we can use it with ease would be as a replacement for trackpads. The current trackpads are quite inefficient and hard to use even with experience. Using a pen as an input device for a track pad could solve that issue since a pen does not need “space” to help the user navigate the screen,

Andrew Menzies 8:26:45 9/23/2014

CrossY: A Crossing-Based Drawing Application Georg Apitz & François Guimbretière In this paper, the authors talk about a new kind of interface for pen-based tablet computers. These interfaces focus on selecting options by using a pen to cross lines rather than tap on buttons. The authors first point out the reasons for developing touch-screen interfaces. Since pens are often used to write and draw, which are a different set of actions than clicking on a specific point, the user does not need to change how they use the device in crossing-based menus. Further, precise clicking can be difficult because the user’s hand gets in the way of their vision, but crossing lines is at least as easy and does not require much more space. (The paper supports this claim by citing work of Accot, who published experiments that involved gate-passing, among other tasks.) The paper also brings up some potential issues with crossing, such as the fact that it can be difficult to cross the correct line when they are all in a row, and offers workarounds such as making the selection lines for items in a vertical checklist diagonal so they can be crossed horizontally (one at a time) or vertically (to choose many in a row). The paper also proposes a few standards for crossing-based interfaces. It mentions, for instance, that people should be able to confirm menu choices by sliding left or down and cancel them by sliding right or up. It also suggests implementing file systems by showing users filename prefixes that the user can select from by swiping up or down and confirm by swiping to the right. Prefixes the user confirms are replaced by steadily longer ones to let the user choose files whose names start with that prefix. While I believe both standards are good ideas, I have some minor concerns about them. One is the idea that menu confirmation and cancellation are always mapped to the same direction. I believe it would be more intuitive if swiping toward where the menu came from cancelled any changes and swiping in any other direction confirmed them. This would agree with the paradigm that the user can undo actions (here, opening the menu) by tracing backwards over the line that triggered them. Also, the idea of navigating a list by expanding prefixes is efficient for large lists where the user already knows what he or she wants. However, it seems inefficient for small lists (where the choices all fit on the screen) and for whenever the user wants to browse through a medium-sized list from top to bottom without a specific item in mind. In those cases, a traditional point-and-click list (or a list where only one line crossing is necessary to choose an item) might be a useful option for the user to be able to turn on. Digital Desk Pierre Wellner This paper presents the idea of a digital desk, where people can work with documents by physically manipulating the paper they are written on. People can perform operations on the computer by placing paper on the desk, moving it around, writing on it, and pointing to it, just as they might do to paper normally. The goal of the digital desk, according to this paper, is to let workers use traditional paper and computers together in an intuitive way. Paper, the work says, has some advantages, such as easy portability, and computers have others, including the ability to calculate quickly. People are used to manipulating both differently. The digital desk, however, reduces the differences between those manipulations. Where a user might move an icon representing a document in a direct manipulation interface, here the user moves the document itself, thereby shrinking the gulf of execution by removing a level of abstraction. This idea presents several challenges. The paper already addressed some of those, saying that optical-character recognition generally succeeded (thanks to local thresholding to separate characters from background), and shadows caused by lights and the users’ hands generally did not cause problems. The paper mentions that the direction people drag their fingers to select a rectangular region matters. Moving in the wrong direction will cause the computer to include the user’s hand in the selection. This may not be a problem because the wrong directions swipe away from the user, but in my experience using mouse-based interfaces, I tend to naturally drag the mouse towards me while selecting a rectangle. One challenge for the users might be discovering what the interface can and cannot do. The paper does not mention how to combine pages into a single document, delete text, or rotate text or images, and I can imagine a few different possible ways to do each. The creators of the system will have to be sure to provide usage hints and feedback to help users discover these operations.

Wenchen Wang 8:45:53 9/23/2014

interacting with paper on the DigitalDesk Summary: This paper introduces a very novel idea that to make physical desk analogous to electronic workstations by taking advantage of desktop metaphor. Since physical tools can be hard to replace, digitaldesk could be more efficient in digital way to become a paperless office. Paper: DigitalDesk is a space where documents coud be both paper and electronic at the same time, so that we could obtain both the strength of physical desk and electronic workstations. By adding the computer to the real work of the user could make user’s work more efficient. Users could copy a long text by just cut off some part of the virtual paper. DigitalDesk could also share editing of documents by continuously grabbing images from local desk and projects scaled, thresholded images from remote desk, so that both users could see what is on both desks. CrossY: A Crossing-Based Drawing Application Summary: This paper explored a crossing as a graphic user interface by taking into account the shape and direction of the strokes. The fluid of crossing could lead to more efficient and natural interfaces for pen computing. Paper review: Nowadays, tablet pc is more and more popular to use. Tablet pc has features, such as portable without a keyboard, direct interaction with low precision. Crossing is invented due to tablet’s appearance. The motivation of crossing is that use of strokes to cross target is more pen friendly. Crossing is as efficient as point-and-click. Crossing is feasible as sole interaction technique and more flexible. It supports command composition and fluid transition between novice and expert. The principle of crossing is that it is based on auto-completion idea, only unique prefixes are presented, selection always at the center.

Yingjie Tang 8:52:49 9/23/2014

The paper Interacting with Paper on the Digitaldesk propose a kind of digital desk on which people can deal with their paper works more like to work on digital works by the use of projector and camera. And with a digitaldesk people can work on a digital work more like to work on a paper work by applying the monitor finger tracking technology. I think some of the views on the paper are useful, for example, users can communicate with each other by editing their paper works or by the gesture. This can inspire the thoughts of a man when he saw the comments appear on his paper at the same time as he thought of it. And the they also propose that people could have a group chat through the camera, this is very important to have a instant group meeting with the group members. However, the reason why digitaldesk has not achieve a lot is obvious. I think this is because their is another way to implement the functions of a digitaldesk: the computers and smartphones. Many of the functions have already been implemented by the personal computer or smartphones. People can use laptops to have a group meeting with the embedded camera and we can scan the paperworks and transfer them to our friends by the smartphone. People are more likely to implement more functions on existing devices rather than to explore a new device. People are likely to write programs for a smartphone or a laptop rather than to work on a plain wooden desk. The paper gives us an experience that when we make design, we should at first choose to take advantages of the existing platforms.———————— The article CrossY: Acrossing-Based Drawing Application mainly introduces the implementation of a stylus interface. The crossing adopted the pen’s natural affordance of drawing strokes, like the color and the degree of thickness. Some of the design method in it are very useful. Users can steering through the goal to replace the point and click function in point-and-click interface. The absolute directory is applied in order to implement the directory browse function, this is very useful. With the absolute directory, users can largely enhance the efficiency of reach the hidden menu. Some designs in the paradigm are in accord with the design principle, the feedback in pointing is very useful, this can not only gives users a feedback but also can agitate the users interests in using this interface. Although the CrossY interface in some level has the advantages overweight the advantages of point-and-click, sometimes it is even more expressive as the standard point-and-click paradigm, personally, I think it can not work very well without a keyboard. The input speed with a keyboard is not comparable by stylus paradigm. With the input software, people can input much more faster than they could write.

Vivek Punjabi 8:59:35 9/23/2014

CrossY: A Crossing-Based Drawing Application The author introduces CrossY as a simple drawing application which uses goal crossing as the basis for graphical user interface. The paper emphasis on the use of crossing in place of point-and-click interface as it is not well adapted to the portable, pen-based computers and tablets. The work demonstrates the feasibility of crossing as an interaction paradigm as a real life application and it also provides initial feedback on the unique challenges of developing such a crossing-based interface. The paper is limited to the exploration of following key concepts: expressiveness, fluid composition of commands, efficiency and visual footprint. It then gives a brief idea about the previous work followed by the detailed uses and advantages of CrossY such as navigation, find and replace, loading drawings, etc. Finally, it compares it with the traditional methods where it results into that CrossY is either similar or better in most of the aspects. This study looks too basic and outdated. As he technology has advanced, new techniques have evolved that addresses these problems in a much better way. So the approach can be modified pertaining to the new technologies and update them regularly. I wasn't yet convinced with the use of crossing as the being the fundamental block and felt that traditional method still looks better. Interacting with paper on the Digital Desk: This paper is about the Digital Desk which has the potential replace the desktops by converting the workstations into Digital Desks rather than the inverse. The author starts with some example applications such as Calculator, Paper Paint and Double Digital Desk. These look pretty cool. Then the implementation issues are explained which includes interaction on the desk, paper reading, thresholding and self calibration. These imply that Digital Desk has some limitations or working conditions that limit its usage. But there are ways to address each of these issues. Then the user experiences are mentioned which provide more usefulness of the idea especially obscuring selections. Finally, the author ends with some good humane quotes which says to us to change the perspective to try to make the computers work in our world rather than the other way round. The author has given a very brief idea about Digital Desk. It doesn't focus much on the technical aspects which I felt was lacking in the paper. Also, the experiences and addressal of the issues could have been expanded. According to me, the main issue lies with the efficiency and accuracy of the product and that should be focused initially. Also, the desk can be made portable rather than being fixed which would add many more applications and uses to the modern day users. And synchronization can be added as a feature to make it more custom.

Eric Gratta 9:23:38 9/23/2014

The Digital Desk (1993) Pierre Wellner This paper was valuable to read not for its proposal of the DigitalDesk but for its perspective on computers: “Do we think of ourselves as working primarily in the computer but with access to physical world functionality, or do we think of ourselves as working primarily in the physical world but with access to computer functionality?” This is a view I’ve never heard or even thought of before. Honestly, I usually assume that the sort of technologies that “bring the computer into our world” are possible but bogged down by practical inefficiency and high cost to implement. To augment the physical world with a computer, the computer either has to have high processing power, many expensive hardware components, and complex algorithms implemented to interpret the real world OR assume a heavily controlled environment and limit the set of experiences possible for users, making the device unattractive and economically infeasible. The DigitalDesk struck me as a fairly novel idea, but by reasoning either intuitively or historically (since, 21 years after that paper, the long nose of innovation is smelling elsewhere), it is an infeasible method of working. Either way, the author made some observations about the limitations of augmenting reality that are useful looking forward. The biggest limitation seemed to be that the entire device depended upon being stationary and calibrated to a single view – of the desk top. A person’s desk space is very personal and variable – they will want to use it for chance storage of items when walking into a room, eating food, reading a book, keeping personal photos and trinkets, etc. It’s not clear from the paper to what extent the user had to make their desk completely clear to support the operation of DigitalDesk, but it seems to interfere with more of the physical world than it augments. The only thing augmented by DigitalDesk is paper, but paper and hands are not the only thing on a desk. That said, I do really wish (2 decades later) there were better alternatives to paper than screens. Rather than dwell on the limitations of DigitalDesk, which the author addressed fairly thoroughly, it would be more useful to think of where a system like this could be adapted. Having a system work in a heavily controlled environment seems pretty important to its feasibility economically. But on a practical level, even if the DigitalDesk made the enhancements that the author addressed, its consumption of the whole desk is inexcusable. There are three directions where I can imagine improvements happening. One direction is to reduce the system-controlled space. The control comes from the limitations of the camera, so either the system as a whole could be reduced in size to just a portion of the desk, or the camera could be given smaller-scale, more specialized functions to reduce possibilities for mobility or calibration (of course, this may alter entirely the vision of the DigitalDesk). Another direction is to convert unused real-world spaces that inherently have little variability into system-controlled spaces. In this way, the system can be more reliable with a more heavily controlled environment, but not interfere with users’ existing personal habits. What are the unused spaces in an office? Are there ones nearby or even on the desk where this system would be most useful and interfere the least? Are there unused spaces surrounding us that might have more utility if they were enhanced by a digital system? The final direction is to put control in the hands of the user. In the case of the DigitalDesk, it feels like the computer is augmenting paper for the user. The user has to follow the specific conventions of the system. But the user should be using DigitalDesk as a tool to augment their papers. How can users be given more control? Can we give them more physical objects to work with and use sensors to detect their position? Can we avoid the ambiguity of digital feedback? This review is already very long, so I can’t write out my ideas for those directions, but this paper was very helpful for sparking my imagination about how to make technology helpful to people. -------------------------------------------------------------------- CrossY: a crossing-based drawing application Georg Apitz “Crossing” will be the fundamental interaction in this application. Author cites evidence that pen-based interactions do not work well or are not easy to perform in a traditional WIMP interface, because the stylus or pen is not as stable as the mouse pointer and is naturally suited for fluid motions, rather than the segmented pointing interactions required by WIMP applications that expect mouse input. Wisely made the decision to try and replicate familiar WIMP functionality in a crossing-based interface, rather than create completely new design interactions, such that the crossing-based interactions could first be accurately evaluated. Reference the steering law from the Accot paper we read, Beyond Fitts’ Law: Models for Trajectory-Based Interactions. Really clever transformation of radio buttons, checkboxes, and OK/Cancel. The color highlighting of the edges makes it clear, and the OK action being in the anticipated direction of the stroke is great design. Concessions: Inconsistency with crossing direction may be confusing, not certain. Tracking the pen/stylus before touching the screen, a feature only available on certain devices, is important for gesture-recognition accuracy. Discoverability of features is difficult; there aren’t many signifiers of the crossing gesture or direction.