In John Searle's paper, "Minds, Brains, and Programs", he discusses the question of whether or not computers have the ability to understand. Searle, through his thought experiment of the Chinese Room, goes on to state that a computer does not in fact have the ability to comprehend or understand, it merely simulates the ability. After reading Searle's paper and reviewing his arguments and counterarguments, I find that I side with him on the matter. In my opinion, the brain is more than just an information processing system, and Searle does a good job of proving that.
In order to prove his point, Searle first describes a computer program that takes in Chinese characters, runs them through a program, and outputs a response in Chinese. Searle's claim is that even if a computer can do this so well that it passes the Turing test, it still does not understand Chinese. The next step in his argument is that of replacing the computer with a human. The human does not speak or understand Chinese at all, but has a set of instructions (i.e. the program) that are in English, that tells him what to do with the input Chinese symbols. As can be deduced from imagining this scenario, even if the person can take in the Chinese input and follow the instructions of outputting proper Chinese symbols, enough so to carry on a conversation with a native Chinese speaker, the person still does not understand Chinese.
With this scenario, it seems easy to be able to differentiate between understanding and merely simulating. The human in the room is merely processing information and not comprehending or realizing what is being communicated. As Searle puts it, the symbols have syntax but no semantics. Just because the person is answering correctly in Chinese, does not mean he knows what he is answering because the symbols mean nothing.
Now, some proponents of "strong AI" (as Searle calls computer AI that understands) will argue that by this thinking, the human brain does not even understand because it is merely a machine that is calculating input and producing output. I think this is where Searle really makes his strongest point. The human brain is something more than just the informational processing. It does not just sort information but comprehends and understands unlike other "machines" can do. His analogy is that we cannot produce sugar and milk from photosynthesis and lactation simulations, so it seems silly that people assume we can produce intentionality from artificial intelligence. There is something natural about the brain that just can not be reproduced with computer hardware.
Wednesday, September 12, 2012
Monday, September 10, 2012
Book Reading #1: Design of Everyday Things
Reaction to the Book in General:
"The Design of Everyday Things" will forever change the way I look at objects and devices and especially doors for my foreseeable future. Donal Norman does a good job of pointing out flaws in designs (maybe too good of a job), and provides a basis for how things should be implemented if we want a more carefree life with our devices. Although it is clearly set in the past, I think his philosophy still applies to design in today's time.
His principles of design hold true because they capture not an issue with a certain device but a way of designing. His principles of design are:
Chapter 1 Reaction:
This book brought about many new ways of thinking for me. I have only read chapter one so far, but it is already changing the way I look at "everyday things." Norman makes many great points about the complexity of technology these days and how the many functions of our devices are making these harder as opposed to easier. I completely agree with this view and find that there is a definite need for more research of design when a product is created. With so many functions crammed into devices, operating them is becoming less and less intuitive. My parents just bought a new Ford Explorer with all of the in dash video controls and light and A/C adjustments, it makes me feel like I'm on a space ship. If it makes me -- the computer engineer -- feel overwhelmed, then I know they don't stand a chance of using all the functions. I don't necessarily blame the designers for these difficulties, because I know I would find it hard to please the manufacturer, the stores, the customer and deciding between functionality and usability; but I think Norman is on to something when he points out this great problem that affects most of our everyday things.
Chapter 2 Reaction:
In chapter 2 of "Design of Everyday Things", Norman discussed the Psychology of everyday actions. He points out that people tend to blame themselves instead of the design of an object when it comes to operating what should be a simple device. I agree with him on this matter and I think the keyboard and secretary example is a great illustration of this problem. The secretaries were inappropriately using the return/enter key on the keyboard, but did not report anything because they assumed it was their fault instead of blaming the design of the keyboard. As Norman points out, a lot of these errors of everyday actions could be fixed with better feedback and visibility built into the designs of everyday things.
Chapter 3 Reaction:
Chaper 3 got a little philosophical on me, and I was worried he would not tie it back into the design of things, but in the end, Norman brought together how Knowledge in the Head and in the World relates to design. As he points out, if something is out of sight, it is out of mind. You can't expect users to remember arbitrary things (rote learning). One good point that I think Norman made is that "if it needs labels, then you should rethink your design because it is flawed." You can give subtle hints with knowledge in the world (such as the stove top switches having mapping that is equal to location of the burners), but expecting the user to have knowledge in the head about all the devices these days is bad design. (Side note: I did not appreciate how he did not say which penny was the correct version. I had to go look it up because I did not know either.)
Chapter 4 Reaction:
In chapter 4, Norman starts to talk about "knowing what to do" when it comes to using devices. I think his viewpoint of making everything more visible and with better feedback, but I think in this chapter he takes it a little too far. I think Norman starts to totally disregard style in his designs and aims completely for functionality instead. I just do not think this is a plausible way of doing things. Everything should not have to be completely self explanatory at the cost of everything else. I think he underestimates people and their ability to deduce how an item is used. Just because some people make a mistake every once in a while, does not mean that a component is a failure.
Chapter 5 Reaction:
Chapter 5 discusses how "to err is Human", which is essentially saying that everyone makes mistakes. Norman discusses how it is the designer's job to make those mistakes less costly. Some good points he makes are putting yourself in the user's point of view and assume that every possible mishap will happen (which is key, because users are not always the smartest when it comes to operating devices). He also talks about making actions reversible, which I think is a great point and also very applicable to CHI and my work in Computer Engineering as a whole. This cannot always be done though, and that is why I think Norman's point of making irreversible errors less costly as a great philosophy.
Chapter 6 Reaction:
I think chapter 6 is one of the chapters that actually starts to point out the difficulties of being a designer. I feel in the earlier chapters so far, Norman has just complained about how hard devices are to use and disregarding style, costs, etc. Chapter 6 references the design challenge and the difficulties of finding the right balance between costs, usability, and style. His take on the computer was very impressive, considering the time written. The amazing thing is that even today, his remarks still apply in the sense that the goal of computing is to make the computer disappear.
Chapter 7 Reaction:
This final chapter of the book was very thoughtful and somewhat tied everything discussed before it together. Chapter 7 focused on user-centered design, which is essentially what the book is all about -- providing a product the user can use efficiently and learn easily. I found the mention of standardization an important topic and I think Norman addressed it well (as he points out, the keyboard and analog clock have always bothered me with their designs). I also liked the section on deliberately making things difficult, because this is something that is not often thought of even though security and other factors he mentions are very prevalent. The ending to his book also drives home the point of his book. He wants to stop not only bad design, but the support of it. He obviously has a lot of peeves when it comes to designs (as he mentions many), and it has a lot to do with the fact that we keep buying the poorly designed products like he says. I will definitely send weeds to the next designer who doesn't practice good design.
5 Examples of Good Design:
Glow in the Dark Light Switch
This light switch is very good design because it's off mode provides both good feedback to the user that it is off as well as a helpful indicator as to where it is. This is great because if a light is off, that means it is usually dark. This was great foresight by the designer for thinking ahead to how the object would be used. This was a very simple but eloquent design feature, not too obtrusive, but jus the right amount of indicator light.
Apple Laptop Charger
The Apple laptop charger is a great device for many reasons and you can definitely tell some serious thought went into what would otherwise seem a trivial device. First off, the plug in is magnetic, so as to aid in the proper placement of the charger; try to plug it into the ethernet port (right by it) and it does not click in. It also has the physical constrains of only fitting two ways, and both ways work. The third reason it is of good design is because of its indicator light. It provides a visual to the user to let them know that the laptop is charging, along with changing from orange to green when fully charged. This way the computer does not even have to be opened to know if it is charging and/or completed charging.
Styrofoam Cooler
This styrofoam cooler is proof that just because something is made to be bought for cheap, does not mean it has to be cheap. Some solid design was put into this simple device, as it has the physical constraint of not opening and spilling while being carried. The handle puts pressure on the lid latches to keep the lid locked. For an object that did not cost a lot, it was designed to prevent error and does a good job of it.
MaroonBike (Chain-less & Air-less)
The maroon bike makers definitely had their users in mind when they were creating their design. With airless tires (solid polyurethane), flats are not an issue. With no chain, there is nothing for your pants to get caught in while riding. It also has a built in bike lock and bike powered LED lights. It just seems everything was thought of when this bike was designed for the casual college student just trying to get around campus, but all these features kept the same operation of a normal bike and very useable.
Helpfully Shaped Key:
This key is a very great design. It solves the problem of keys looking ambiguous and yet only fitting one way. With the black portion, it affords a grip like a pistol, and therefore the correct orientation for insertion into the ignition. This is very helpful at night when scrambling for the key and being able to know that it correctly held for proper placement.
5 Examples of Bad Design:
Ceiling Fan
Which String Controls the Light...and Which Fan Speed am I On?
I think all ceiling fans have this issue and it has tricked me many times over the years. The draw strings seem identical and it is very hard to tell which controls the light and which controls the fan. I have on many occasions pulled the fan string hoping to turn off the lights. You then have to pull the fan string multiple times until you get back to the proper speed. That brings me to my second issue with the ceiling fan -- there is no indication of which fan speed you are currently on -- no feedback. You have to just wait for the fan to either slow down to make sure it is off or speed all the way up and guess by the sound of it that it is at full speed. Poor design that seems to go unnoticed.
Fishing Lure Box
Don't forget to lock before lifting handle...
The fishing lure box is a great invention. You can store all sorts of different lures and hooks and bobbers etc. in it. It conveniently pops out so you can view all of these easily. My complaint is that the handle is on the lid. You may only do this once, because it is so frustrating, but if you ever forget to latch the front lock on the box before picking up by the handle, the lid pops open and spits out everything. The trays are supposed to easily slide out, and so when given this opportunity, they will gladly pop out and spill all contents inside. So while the inside of the box is a great idea, location of the handle could maybe be reconsidered.
Surge Protector
This surge protector has a couple problems in my mind. First off, the 3-prong plugs afford you to plug in your devices vertically. This is great except for when you have a full protector with multiple 3-prong plugs because these types of plugs generally then take up two sockets because they cover up the second one with where the cord is trailing (or in the case of the Apple charger, the box covers up a socket). I do not know if this is the fault of the surge protector or the plug designers because now there are plugs that point out sideways, and so there now seems to be no fully compatible way to orient the sockets on the surge protector. This seems to be an issue of no standardization. A second problem with this surge protector's design is that the on/off switch is on the same side as the sockets. This gives way to the problem (as illustrated in the picture) of larger plugs completely blocking the on/off switch. It seems this popular version of the surge protector could use some better design.
The toaster oven has a two features that I think merit bad design. The first problem is the timer (bottom right knob). In order to operate on times longer than 10 minutes, you just turn the knob to the choice, but to operate for less than 10 minutes, you must first turn the knob to 10 minutes and then back to the lesser value. Nowhere does the toaster indicate that this would be necessary. It just provides a poor conceptual model to the user because no one would think to turn past and then back to the time they want. Also, the feature of having to reach into a small hot oven to get out your food is poor design as well. A possible solution would be to have the oven tray slide out with the door when it is opened so there would be less burning of the top of the hand.
The Wii-Mote = Broken TV Screen
The wii controller is an example of bad design because of it's smooth grip. The designers were obviously focusing more on style instead of thinking about how the device would be used. The smoothness of the controller creates the issue of it flying out of the user's hand because of the rapid movement involved in wii gaming. Yes, Nintendo did try to solve this by adding the strap, but as can be seen in the pictures, it breaks quite easily due to it's thin nature (and that is if the user even takes the time to put it on). This could have been solved with a simple rubber grip, but the designers did not design with error in mind, but instead with style and looks in mind.
"The Design of Everyday Things" will forever change the way I look at objects and devices and especially doors for my foreseeable future. Donal Norman does a good job of pointing out flaws in designs (maybe too good of a job), and provides a basis for how things should be implemented if we want a more carefree life with our devices. Although it is clearly set in the past, I think his philosophy still applies to design in today's time.
His principles of design hold true because they capture not an issue with a certain device but a way of designing. His principles of design are:
- use both knowledge in the world and knowledge in the head
- simplify the structure of the task
- make things visible
- get mappings right
- exploit the power of constraints (natural and artificial)
- design for error
- standardize if all else fails
When you look at that list, you realize that if it was used more frequently in production of today's devices, life would be a bit simpler.
The three principles that really stood out to me from this list, though, were "making things visible", "getting mappings right", and "designing for error." Making things visible is so important because, without providing feedback to the user, there is no way for them to know if what they are trying to do is actually being done. A design can only be successful if the the gulf of execution and evaluation is properly bridged.
The freezer/refrigerator model of a poor system image that Norman gives is really a great illustration of a system image gone wrong and therefore a poor conceptual model for the user. This can be seen as poor visibility for the user, or as my second favorite point is, it could be poor mapping. I think getting mapping right is so very important and there is often a struggle with that with devices these days due to the "creeping featurism"(the tendency to add to the number of features a device can do). It is getting harder and harder to provide the functions users want, but also make it where they actually know how to use them. I think the dilemma of "more buttons means more complexity", but "less buttons means poor mapping" is a tough one to figure out, but it is just something that designing with the user in mind entails.
This idea of designing for the user isn't more relevant than with the issue of designing for error. People are going to make mistakes with devices and use them the wrong way, no matter how well you think you designed it. Designing with this in mind though, as Norman states, makes the device much more useable. By making actions easy to undo or permanent actions harder to do, lots of frustration can be avoided.
Overall, I think the book was a very good read. At points I did get frustrated with what seemed unnecessary complaining from Dr. Norman, as it seemed he exaggerated the problems of objects to prove his point (i.e. the light switches and faucet examples). I do think his principles of design are excellent as has been proven by their stand of the test of time, and, as I said, I will definitely appreciate the design of everyday things a lot more now.
Chapter 1 Reaction:
This book brought about many new ways of thinking for me. I have only read chapter one so far, but it is already changing the way I look at "everyday things." Norman makes many great points about the complexity of technology these days and how the many functions of our devices are making these harder as opposed to easier. I completely agree with this view and find that there is a definite need for more research of design when a product is created. With so many functions crammed into devices, operating them is becoming less and less intuitive. My parents just bought a new Ford Explorer with all of the in dash video controls and light and A/C adjustments, it makes me feel like I'm on a space ship. If it makes me -- the computer engineer -- feel overwhelmed, then I know they don't stand a chance of using all the functions. I don't necessarily blame the designers for these difficulties, because I know I would find it hard to please the manufacturer, the stores, the customer and deciding between functionality and usability; but I think Norman is on to something when he points out this great problem that affects most of our everyday things.
Chapter 2 Reaction:
In chapter 2 of "Design of Everyday Things", Norman discussed the Psychology of everyday actions. He points out that people tend to blame themselves instead of the design of an object when it comes to operating what should be a simple device. I agree with him on this matter and I think the keyboard and secretary example is a great illustration of this problem. The secretaries were inappropriately using the return/enter key on the keyboard, but did not report anything because they assumed it was their fault instead of blaming the design of the keyboard. As Norman points out, a lot of these errors of everyday actions could be fixed with better feedback and visibility built into the designs of everyday things.
Chapter 3 Reaction:
Chaper 3 got a little philosophical on me, and I was worried he would not tie it back into the design of things, but in the end, Norman brought together how Knowledge in the Head and in the World relates to design. As he points out, if something is out of sight, it is out of mind. You can't expect users to remember arbitrary things (rote learning). One good point that I think Norman made is that "if it needs labels, then you should rethink your design because it is flawed." You can give subtle hints with knowledge in the world (such as the stove top switches having mapping that is equal to location of the burners), but expecting the user to have knowledge in the head about all the devices these days is bad design. (Side note: I did not appreciate how he did not say which penny was the correct version. I had to go look it up because I did not know either.)
Chapter 4 Reaction:
In chapter 4, Norman starts to talk about "knowing what to do" when it comes to using devices. I think his viewpoint of making everything more visible and with better feedback, but I think in this chapter he takes it a little too far. I think Norman starts to totally disregard style in his designs and aims completely for functionality instead. I just do not think this is a plausible way of doing things. Everything should not have to be completely self explanatory at the cost of everything else. I think he underestimates people and their ability to deduce how an item is used. Just because some people make a mistake every once in a while, does not mean that a component is a failure.
Chapter 5 Reaction:
Chapter 5 discusses how "to err is Human", which is essentially saying that everyone makes mistakes. Norman discusses how it is the designer's job to make those mistakes less costly. Some good points he makes are putting yourself in the user's point of view and assume that every possible mishap will happen (which is key, because users are not always the smartest when it comes to operating devices). He also talks about making actions reversible, which I think is a great point and also very applicable to CHI and my work in Computer Engineering as a whole. This cannot always be done though, and that is why I think Norman's point of making irreversible errors less costly as a great philosophy.
Chapter 6 Reaction:
I think chapter 6 is one of the chapters that actually starts to point out the difficulties of being a designer. I feel in the earlier chapters so far, Norman has just complained about how hard devices are to use and disregarding style, costs, etc. Chapter 6 references the design challenge and the difficulties of finding the right balance between costs, usability, and style. His take on the computer was very impressive, considering the time written. The amazing thing is that even today, his remarks still apply in the sense that the goal of computing is to make the computer disappear.
Chapter 7 Reaction:
This final chapter of the book was very thoughtful and somewhat tied everything discussed before it together. Chapter 7 focused on user-centered design, which is essentially what the book is all about -- providing a product the user can use efficiently and learn easily. I found the mention of standardization an important topic and I think Norman addressed it well (as he points out, the keyboard and analog clock have always bothered me with their designs). I also liked the section on deliberately making things difficult, because this is something that is not often thought of even though security and other factors he mentions are very prevalent. The ending to his book also drives home the point of his book. He wants to stop not only bad design, but the support of it. He obviously has a lot of peeves when it comes to designs (as he mentions many), and it has a lot to do with the fact that we keep buying the poorly designed products like he says. I will definitely send weeds to the next designer who doesn't practice good design.
5 Examples of Good Design:
Glow in the Dark Light Switch
This light switch is very good design because it's off mode provides both good feedback to the user that it is off as well as a helpful indicator as to where it is. This is great because if a light is off, that means it is usually dark. This was great foresight by the designer for thinking ahead to how the object would be used. This was a very simple but eloquent design feature, not too obtrusive, but jus the right amount of indicator light.Apple Laptop Charger
This styrofoam cooler is proof that just because something is made to be bought for cheap, does not mean it has to be cheap. Some solid design was put into this simple device, as it has the physical constraint of not opening and spilling while being carried. The handle puts pressure on the lid latches to keep the lid locked. For an object that did not cost a lot, it was designed to prevent error and does a good job of it.
MaroonBike (Chain-less & Air-less)
Helpfully Shaped Key:
5 Examples of Bad Design:
Ceiling Fan
Which String Controls the Light...and Which Fan Speed am I On?
Fishing Lure Box
Don't forget to lock before lifting handle...
The fishing lure box is a great invention. You can store all sorts of different lures and hooks and bobbers etc. in it. It conveniently pops out so you can view all of these easily. My complaint is that the handle is on the lid. You may only do this once, because it is so frustrating, but if you ever forget to latch the front lock on the box before picking up by the handle, the lid pops open and spits out everything. The trays are supposed to easily slide out, and so when given this opportunity, they will gladly pop out and spill all contents inside. So while the inside of the box is a great idea, location of the handle could maybe be reconsidered.Surge Protector
Toaster Oven
The toaster oven has a two features that I think merit bad design. The first problem is the timer (bottom right knob). In order to operate on times longer than 10 minutes, you just turn the knob to the choice, but to operate for less than 10 minutes, you must first turn the knob to 10 minutes and then back to the lesser value. Nowhere does the toaster indicate that this would be necessary. It just provides a poor conceptual model to the user because no one would think to turn past and then back to the time they want. Also, the feature of having to reach into a small hot oven to get out your food is poor design as well. A possible solution would be to have the oven tray slide out with the door when it is opened so there would be less burning of the top of the hand.The Wii-Mote = Broken TV Screen
The wii controller is an example of bad design because of it's smooth grip. The designers were obviously focusing more on style instead of thinking about how the device would be used. The smoothness of the controller creates the issue of it flying out of the user's hand because of the rapid movement involved in wii gaming. Yes, Nintendo did try to solve this by adding the strap, but as can be seen in the pictures, it breaks quite easily due to it's thin nature (and that is if the user even takes the time to put it on). This could have been solved with a simple rubber grip, but the designers did not design with error in mind, but instead with style and looks in mind.Sunday, September 9, 2012
Paper Reading #6: ZeroN: mid-air tangible interaction enabled by computer controlled magnetic levitation
Paper Reading #6: ZeroN: mid-air tangible interaction enabled by computer controlled magnetic levitation
Intro:
- Title - ZeroN: mid-air tangible interaction enabled by computer controlled magnetic levitation
- Reference Information:
ACM Classification: H5.2 [Information interfaces and presentation]: User Interfaces.
General terms: Design, Human Factors
Keywords: Tangible Interfaces, 3D UI.
- Author Bios - Jinha Lee, Rehmi Post, and Hiroshi Ishii
Hiroshi ishii is a Professor at MIT and leader of the Tangible Media Group, which explores the tangible bits vision to seamlessly couple the dual world of bits and atoms by giving physical form to digital information. Also, has an impressive publication count of 154, dating back to 1990.
Jihna Lee was a graduate research assistant for the Tangible Media Group, and now works at the MIT Media Laboratory.
Rehmi Post works at the MIT Center for Bits and Atoms. He has 4 publications dating from 1994 to 2012.
Summary:
The work of this paper is called ZeroN, a new tangible interface element that can be levitated and moved freely by a computer in a three dimensional space. ZeroN serves as a tangible representation of a 3D coordinate of the virtual world through which users can se, feel, and control computation. This is done by using a magnetic control system that can levitate and actuate a permanent magnet in a predefined 3D volume. There is also optical tracking and display system that projects images on the levitating object.
Related work not referenced in the paper:
- Emerging frameworks for tangible user interfaces - B. Ullmer, H. Ishii
- The reacTable: exploring the synergy between live music performance and tabletop tangible interfaces - Sergi Jorda, Gunter Geiger, Marcos Alonso, Martin Kalenbrunner
- Tangible interfaces for remote collaboration and communication - Scott Brave, Hiroshi Ishii, Andrew Dahley
- A taxonomy for and analysis of tangible interfaces - Kenneth P. Fishkin
- Classroom collaboration in the design of tangible interfaces for storytelling - Danae Stanton, Victor Bayon, Helen Neale, Ahmed Ghali, Steve Benford
- The actuated workbench: computer-controlled actuation in tabletop tangible interfaces - Gian Pangaro, Dan Mayes-Aminzade, Hiroshi Ishii
- Extending tangible interfaces for education: digital montessori-inspired manipulatives - Oren Zuckerman, Saeed Arida, Mitchell Rescind
- Do tangible interfaces enhance learning? - Paul Marshall
- Tangible Query Interfaces: Physically constrained tokens for manipulating database queries - B. Ullmer, H. Ishhii, R. Jacob
- 3D user interfaces : theory and practice - Doug A. Bowman, Joseph J. Laviola, Ernst Kruijff
Evaluation:
The authors first evaluated the work of their project by having a user study and getting feedback on how the system felt and operated. This gave some comments and feelings on possible improvements, as well as praises (qualitative, subjective). There was also a technical evaluation of the system. Measurements of the max levitation height and speed of actuation (quantitative, objective). There was also study on the different degrees of freedom and the resolution of the 3D display. Overall, this was quite a systemic evaluation and good test of if the system works or not. This evaluation was appropriate for the context of the work and what the authors were trying to prove. They found the flaws of the system and what the users liked as well.
Discussion:
I think this project is a very interesting work. It has a lot of possibilities for it's future and I think if it can be continued to be improved with better feedback and larger scale implementations, it can be a very useful tool for the audience it is aiming for. The evaluation was appropriate, as this design is still early in the stages. I do think this is a novel idea. There is some related work, but none that does it how the authors in this work do. Finally, I think the pong game would be awesome if it was the size of a room.
Wednesday, September 5, 2012
Paper Reading #5: KinectFusion: real-time 3D reconstruction and interaction using a moving depth camera
Paper Reading #5: KinectFusion: real-time 3D reconstruction and interaction using a moving depth camera
- Title - KinectFusion: real-time 3D reconstruction and interaction using a moving depth camera
- Reference Information -ACM Classification: H5.2 [Information Interfaces and Presentation]: User Interfaces. I4.5 [Image Processing and Computer Vision]: Reconstruction. I3.7 [Computer Graphics]: Three-Dimensional Graphics and Realism.
General terms: Algorithms, Design, Human Factors. Keywords: 3D, GPU, Surface Reconstruction, Tracking, Depth Cameras, AR, Physics, Geometry-Aware Interactions
- Author Bios - Shahram Izadi, David Kim, Otmar Hilliges, David Molyneaux, Richard Newcombe, Pushmeet Kohli, Jamie Shotton, Steve Hodges, Dustin Freeman, Andrew Davison, Andrew Fitzgibbon
This group works together mainly at Microsoft Research in Cambridge, Cambridge, United Kingdom. Newcombe and Davison are from Imperial College London in London, United Kingdom though, and Freeman is from the University of Toronto in Canada. They all work in the research of input sensor devices, display technologies, computer graphics, and computer human interaction.
Summary:
KinectFusion enables a user holding and moving a standard
Kinect camera to rapidly create detailed 3D reconstructions
of an indoor scene. Only the depth data from Kinect is used
to track the 3D pose of the sensor and reconstruct, geomet-
rically precise, 3D models of the physical scene in real-time. Novel extensions to the core GPU pipeline demonstrate object segmentation and user interaction directly in front of the sensor,
without degrading camera tracking or reconstruction. These
extensions are used to enable real-time multi-touch interactions anywhere, allowing any planar or non-planar reconstructed physical surface to be appropriated for touch.
Poisson surface reconstruction - Michael Kazhdan, Matthew Bolitho, Hughues Hoppe
- Towards Urban 3D Reconstruction from Video - Frahm, J.-M. ; Mordohai, P. ; Clipp, B. ; Engels, C. ; Gallup, D. ; Merrell, P. ; Phelps, M. ; Sinha, S. ; Talton, B. ; Wang, L.; Yang, Q. ; Stewenius, H. ; Yang, R. ; Welch, G. ; Towles, H.; Nister, D. ; Pollefeys, M.
- High-quality surface splatting on today's GPUs - Botsch, M., Hornung, A. ; Zwicker, M. ; Kobbelt, L.
- Data-Parallel Octrees for Surface Reconstruction - Kun Zhou, Minmin Gong ; Xin Huang ; Baining Guo
- Parallel Tracking and Mapping for Small AR Workspaces - G. Klein, D. Murray
- Real-time vision-based camera tracking for augmented reality applications - Dieter Koller, Gudrun Klinker, Eric Rose, David Breen, Ross Whitaker
- Scene modelling, recognition and tracking with invariant image features - I. Skrypnky, D.G. Lowe
A muscle model for animation three-dimensional facial expression - Keith Waters
- An image-based approach to three-dimensional computer graphics - Leonard McMillan Jr.
- Merging virtual objects with the real world: seeing ultrasound imagery within the patient - Michael Bajura, Henry Fuchs, Ryutarou Ohbuchi
These areas of study -- active sensors, passive cameras, unordered 3D points, etc -- are highly studied ares of research in computer graphics and vision, but KinectFusion differentiates itself by doing things better with interactive rates, no explicit feature detection, high-quality construction of geometry, and dynamic interaction. KinectFusion is also infrastructure-less with whole room reconstruction and interaction.
Evaluation for this project was not very systemic. There were no quantitative or subjective evaluation measures done. There was no actual evaluation process, per se, in the paper, but it was more of how they did it and proof that it worked as they said. There were tests of each of the features of KinectFusion run, but there were no hard results other than whether it worked and what problems they ran into. This can be seen as qualitative, and measuring it's completeness. This is appropriate though, as there is no other applicable way to test this work other than to have users try it out and give their subjective opinions.
Evaluation:
Evaluation for this project was not very systemic. There were no quantitative or subjective evaluation measures done. There was no actual evaluation process, per se, in the paper, but it was more of how they did it and proof that it worked as they said. There were tests of each of the features of KinectFusion run, but there were no hard results other than whether it worked and what problems they ran into. This can be seen as qualitative, and measuring it's completeness. This is appropriate though, as there is no other applicable way to test this work other than to have users try it out and give their subjective opinions.
Discussion:
I think this project work is quite well done. The uses are quite broad as there are many functions to KinectFusion. I thought the physics interaction was the best part, as interacting with a 3D generated room is unique and it is novel in the way they did it. The evaluation could have been a bit more in depth, but they were more in the testing stages of proving it worked and preparing it then creating a prototype that could be tested for specific functions. Overall, I think it is a very worthy contribution and it will be interesting to see where it goes from here.
Paper Reading #4: Personalized input: improving ten-finger touchscreen typing through automatic adaptation
Paper Reading #4: Personalized input: improving ten-finger touchscreen typing through automatic adaptation
Intro:
- Title - Personalized input: improving ten-finger touchscreen typing through automatic adaptation
- Reference Information -
Author Keywords
Touchscreen text input, personalization, adaptive interfaces.
ACM Classification Keywords
H5.2 [Information interfaces and presentation]: User interfaces—input devices and strategies.
Presented at CHI '12 - Author Bios - Leah Findlater and Jacob O. Wobbrock
Leah Findlater - Attends the College of Information Studies at the University of Maryland in College Park, Maryland.
Jaccob Wobbrock - He is from the University of Washington in Seattle, Washington. He is part of The Information School and the DUB Group there.
This is both of their's first paper published in the ACM. As they are both just Graduate students, they are fairly new to the world of CHI.
This work focused on introducing and evaluating two novel personalized keyboard interfaced, both of which adapt their underlying key-press classification models. One of the keyboards even adapted the location of the keys visually. Since personalized
keyboards are still an emerging area of research, they also
outline a design space that includes dimensions of
adaptation and key-press classification features. Because of their evaluation, they determine pros and cons of different modes of touchscreen typing through automatic adaptation.
Related work not referenced in the paper:
Evaluation:
The evaluation process was a controlled three-session study of 12 participants with both the adaptive and non-adaptive personalized keyboards to a conventional touch keyboard. The main quantitative measurements were speed and uncorrected error rate. The words per minute were calculated for each model and this gave a objective quantitative evaluation. Participants were also asked to rank the three keyboards based
on ease of use, efficiency, frustration, comfort, how natural
the typing felt, and overall preference. This gave a subjective opinion on the work. This gave a systemic evaluation of the touchscreen, as it was tested on efficiency and likability for if it is practical.
Discussion:
- High precision touchscreens: design strategies and comparisons with a mouse - Andrew Sears, Ben Schneiderman
- ThumbSpace: Generalized One-Handed Input for Touchscreen-Based Mobile Devices - Amy K. Karlson and Benjamin B. Bederson
- Pressure-based text entry for mobile devices - Stephen A. Brewster, Michael Hughes
- Adaptive interfaces for ubiquitous web access - Daniel Billsus, Clifford Brunk, Craig Evans, Brian Gladish
- Adaptive Interfaces and Agents - A. Jameson
- Machine learning for adaptive user interfaces - Pat Langley
- Exploring the design space for adaptive graphical user interfaces - Daniel S. Weld, Desney Tan, Mary Czerwinski
- Supporting adaptive interfaces in a knowledge-based user interface environment - James D. Foley, Piyawadee Noi Sukaviriya
- Experience with adaptive interfaces - D. Benyon, D. Murray
- An empirical appraisal of the effectiveness of adaptive interfaces for instructional systems - John Eklund, Ken Sinclair
Discussion:
I think this work is very impressive. Improvement of keyboard is something that could use some work, especially with touch devices. I think their method has promise and should be continued to be evaluated, as it could definitely benefit users during this rise of touchscreen devices. I do not think it was necessarily a novel idea, because it is mainly just improving the keyboard. I do think it is important though and, as the results show, could definitely improve the touch keyboards. The evaluation was appropriate, and I think it was one of the more important parts of this project, as it showed some pros and cons of visual adaptive keyboards.
Monday, September 3, 2012
Paper Reading #3: LightGuide: Projected Visualizations for Hand Movement Guidance
Paper Reading #3: LightGuide: Projected Visualizations for Hand Movement Guidance
- Title - LightGuide: Projected Visualizations for Hand Movement Guidance
- Reference Information -
Session: Curves & Mirages: Gestures & Interaction with Nonplanar Surfaces
May 5–10, 2012, Austin, Texas, USA
Author KeywordsOn-demand interfaces; on-body computing; appropriated surfaces; tracking; spatial augmented reality;
ACM Classification Keywords
H.5.2 [Information interfaces and presentation]: User Interfaces - Input devices & strategies;
- Author Bios - Rajinder Sodhi, Hrvoje Benko, Andrew D. Wilson
 |
| Hrvoje Benko |
Hrvoje Benko - He is a researcher in Natural Interaction Research group at Microsoft Research. His research is on novel surface computing technologies and their impact on human-computer interaction. Prior to working at Microsoft, he received his PhD at Columbia University, working on augmented reality projects that combine immersive experiences with interactive tabletops. His projects explore the space where the flat digital world of surface computing meets the curved, physical, 3D space we live in.
Andrew D. Wilson - Also works at Microsoft Research. He has an impressive 55 publications (1,034 citations) from 1995-2012, so is the veteran of the team.
Rajinder Sodhi - He is a PhD student in Computer Science at the University of Illinois. He works at the intersection of computer vision and human computer interaction and is advised by David Forsyth and Brian Bailey. His research focuses on new display and interaction techniques - especially those that allow people to embed interactivity in all aspects of the environment.
Summary:
LightGuide is a system that uses a projection of hints on a person's body to aid in gesture guidance. These projected hints guide the user in completing the desired motion with their body part which is particularly useful for performing movements that require accuracy and proper technique, such as during exercise or physical therapy. This was done with a low cost depth camera and a projector to display the informational interaction techniques. They then quantified the performance of LightGuide with a user study comparing the use of LightGuide to traditional gesture guide videos. Movements were an astounding 85% more accurate with the use of LightGuide aiding the user.
Related work not referenced in the paper:
- Computer-based training in two-dimensional echocardiography using an echocardiography simulator -
- Michael Weidenbach,
- Florentine Wild,
- Kathrin Scheer,
- Gerhard Muth
- ,
- Stefan Kreutter,
- Gernoth Grunst,
- Thomas Berlage,
- Peter Schneider
- OmniTouch: wearable multitouch interaction everywhere - Chris Harrison, Hrvoje Benko, Andrew Wilson
- Simulating educational physical experiments in augmented reality - Hannes Kaufmann, Bernd Meyer
- AUGMENTED REALITY PLATFORMS FOR VIRTUAL FITTING ROOMS - Ioannis Pachoulakis and Kostas Kapetanakis
- Pfinder: real-time tracking of the human body - Wren, C.R. ; Darrell, T. ; Pentland, A.P.
- Spatially Augmented Reality - Ramesh Raskar, Greg Welch, Henry Fuchs
- Augmented reality visualization for laparoscopic surgery - Henry Fuchs, Mark A. Livingston, Ramesh Raskar, D’nardo Colucci, Kurtis Keller, Andrei State, Jessica R. Crawford, Paul Rademacher, Samuel H. Drake and Anthony A. Meyer
- Combining multiple depth cameras and projectors for interactions on, above and between surfaces - Andrew Wilson, Hrvoje Benko
- A Survey of Augmented Reality - Ronald T. Azuma
- The Everywhere Displays Projector: A Device to Create Ubiquitous Graphical Interfaces - Claudio Pinhanez
The work of this project is in 4 main areas: computer-aided task guidance, task-guidance in augmented reality, augmenting environments with projectors, and projection-based guidance. LightGuide bases off of related work in these areas, such as Feiner et. al using augmented reality to help with laser printer assembly, or support systems for playing guitar or manufacturing. These involved cumbersome head units or other user attachments. There is past work with depth sensing cameras and stationary projectors like LightGuide as well though. There is also former projection-based guidance systems that have been implemented, like LightGuide. LightGuide is different though, because it provides the hints with real-time feedback directly on the user's hand that is tracked in mid-air for guidance. So, LightGuide is novel in the sense that it combines all these features in a way that has not been done before. It provides dynamic feedback, is mid-air and free, and is with a depth camera and projector. The authors appropriately talked about all the other related works and also explained why the aforementioned reasons made theirs unique.
Evaluation:
The authors evaluated their work by running a test on 10 users. The test was to see if the system worked better than the typical gesture guidance system of watching a how to video. The test participants had to trace 5 paths: a line, a square, a circle, the letter 'N", and a line plus curve, all run at different depths to ensure adequately tested variety of depths. Each participant did 90 trials (6 conditions x 5 paths x and 3 angles), which gave quantitative data on how accurately the user followed the given path by time, position, and hand-orientation (quantitative, objective). The users were also interviews after each session and after the complete study to get their opinion on the different methods of hints and the overall effectiveness of the system (qualitative, subjective). This study measure the system as a whole and was very complete and thorough in its evaluation.
 |
| Error Report by Guidance System |
Discussion:
I think that the LightGuide is an interesting project. I understand that it has some possible uses, but I think it still very early in its progress. There was no real-life demo available, just proposed uses. Of these, I think physical therapy the most probable future use. I think the evaluation was appropriate and covered the system quite completely. It got both numbers and opinions on the device to get a full evaluation. I think the contribution is not necessarily novel, but a nice improvement on other systems. The real-time feedback in the LightGuide is what really sets it apart, and I think it has a chance to become something useful in the future.
Paper Reading #2: MirageTable: Freehand Interaction on a Projected Augmented Reality Tabletop
Paper Reading #2: MirageTable: Freehand Interaction on a Projected
Augmented Reality Tabletop
Intro:
- Title - MirageTable: Freehand Interaction on a Projected Augmented Reality Tabletop
- Reference Information -
ACM Classification: H.5.2 [Information Interfaces and Presentation]: User Interfaces - Graphical user interfaces
Session: Curves & Mirages: Gestures & Interaction with Nonplanar Surfaces
May 5–10, 2012, Austin, Texas, USA
Microsoft Research - Redmond, Washington
VIMMI / Inesc-ID IST / Technical University of Lisbon - Lisbon, Portugal
- Author Bios - Hrvoje Benko, Ricardo Jota, and Andrew D. Wilson
 |
| Hrvoje Benko |
Hrvoje Benko - He is a researcher in Natural Interaction Research group at Microsoft Research. His research is on novel surface computing technologies and their impact on human-computer interaction. Prior to working at Microsoft, he received his PhD at Columbia University, working on augmented reality projects that combine immersive experiences with interactive tabletops. His projects explore the space where the flat digital world of surface computing meets the curved, physical, 3D space we live in.
Andrew D. Wilson - Also works at Microsoft Research. He has an impressive 55 publications (1,034 citations) from 1995-2012, so is the veteran of the team.
Ricardo Jota - Works at Microsoft Research, but also at VIMMI / Inesc-ID IST / Technical University of Lisbon in Lisbon, Portugal. He has 10 publications from 2003-2012.
Summary:
MirageTable is a curved projection-based augmented reality system, which digitizes any object on the surface,
presenting correct perspective views accounting for real objects and supporting freehand physics-based interactions. (YouTube link). It focuses on the input of users, instead of the typical augmented reality setups these days that are output oriented. The user can interact with objects without having to wear any additional trackers, gloves, or gear. By simply placing an object on the screen, you can copy it and have a digital copy to interact with. The user's head is also tracked so that the 3D perspective is always correct. Freehand interaction with these virtual objects is much the same way as interaction with their real world object counterparts.
Related Work Not Referenced in the Paper:
- Projector-Based Location Discovery and Tracking -Johnny Chung Lee
- Spatial Augmented Reality - Oliver Bimber, Ramesh Raskar
- Marker tracking and HMD calibration for a video-based augmented reality conferencing system - Kato, H. Fac. of Inf. Sci., Hiroshima City Univ. Billinghurst, M.
- The Visual Computing of Projector-Camera Systems -
- Oliver Bimber,
- Daisuke Iwai
- ,
- Gordon Wetzstein,
- Anselm Grundhöfer
- Making one object look like another: controlling appearance using a projector-camera system - Peri, H.; Nayar, S.K.; Belhumeur, P.N.
- iLamps: geometrically aware and self-configuring projectors - Ramesh Raskar, Jeroen van Baar, Paul Beardsley, Thomas Willwacher, Srinivas Rao, and Clifton Forlines
- Embedding imperceptible patterns into projected images for simultaneous acquisition and display - Cotting, D.;Naef, M. ; Gross, M. ; Fuchs, H.
- A Multi-Camera Method for 3D Digitization of Dynamic, Real-World Events - Peter Rander
- Real-Time 3D Model Acquisition - Szymon Rusinkiewicz, Olaf Hall-Holt, Marc Levoy
- Telepresence: Integrating Shared Task and Person Spaces - William A. S. Buxton
The work these authors did has been demonstrated independently in the past (i.e. the office of the future, the perceptive workbench, depth sensing video cameras), but MirageTable demonstrates how integrating them together creates a compelling augmented reality experience. MirageTable uses these other works to create 3 new advances to the area of study.
- system design and implementation
- three prototype applications (i.e. 3D video conferencing)
- user study on 3D perception and image quality
So, I would not say this work is novel, but it is quite an advancement on previous works in the similar area. The authors did do a good job of pointing out their related work as well, as they reference most similar work in their areas of projection based augmented reality solutions and the use of depth sensing cameras for input.
Evaluation:
The authors' main evaluations of this system were done with projective texturing quality. The first evaluation was of the effect of projection surface on image quality. This evaluation involved measuring the perceived distance of a virtual object when it was projected on varying surfaces -- both color and depth. This was quantified using the Root Mean Square difference between the image on a white background (base) to the image on the varying backgrounds (quantitative, objective). The second evaluation was of effect of projection surface on depth perception. This study was done by having users rate the depth of a sphere floating above the table of various surfaces and colors. This gave a quantitative but subjective evaluation of the project. The users also ranked their most preferred surface to work on for another quantitative subjective evaluation. These evaluations mainly focused on the projective texturing quality and so did not test the whole system of its usability, but their evaluation was a very important part of research in the area of work.
Discussion:
I think the work is a very important step in the advancement of projected augmented reality systems. It does a good job of combining and advancing previous works in the area. I think it does a good job of evaluating the system in the areas that the authors were unsure about. The system as a whole was not tested for user experience and usability, but the area of evaluation was crucial for the system to work. I do not think this was a novel idea in that it is based on combining previous works, but it is a very great advancement. The practical use is hard to see at this point in most of its functions, but the 3D video conferencing is a very impressive idea. The idea of sharing virtual objects on a virtual desk is quite influential, and I see this being something this project creates.
Subscribe to:
Posts (Atom)