Pages

Wednesday, September 12, 2012

Assignment 3 - Chinese Room Blog


Personally, I do not believe that machines can “understand” even if they can produce results that simulate understanding. I believe that anything that does not have a brain, made of tissue and functions through physical and chemical processes, cannot possibly “understand”. There are many complicated processes that make up the human brain and it would be nearly impossible, if not impossible, to completely recreate these processes to create a computer/robot that “understand”. Computers cannot “understand” because calculations and algorithms are used to come up with output, based on input. And, although I believe only humans and some creatures (mammals, birds, other semi-intelligent creatures) are capable of understanding, it could be argued that humans produce output by performing many complex algorithms and calculations in their brain, much like a computer does. From a young age, humans begin to develop “understanding” for a certain language. However, it could be said that as they are growing they are simply learning new rules and algorithms to determine how to answer questions. In a way, we are taking input, performing lots of calculations in our head, and then producing output. To me, “understanding” requires much more than rules and calculations. A mind, along with emotions, is needed to fully understand something. From my last statement, you could guess that I do not agree with “brains cause minds” and “actual human mental phenomena [are] dependent on actual physical-chemical properties of actual human brains”. These statements suggest that what I am doing right now is based on the laws of physics and some chemical processes. While I do not disagree with the fact that there are chemical processes and physics laws that are going as I type this right now, I do disagree with the idea that the mind and emotions are based on these laws. If the brain/mind were driven by physical and chemical processes, then free will would not be possible. Everything that happened in your life or that you have thought of would have been driven by complex physics and chemistry equations. 

I am currently taking an artificial intelligence course, and while I think it is a cool and interesting topic, I believe computers will never be able to “understand” the way a human does. Humans barely “understand” the inner machinations of the human mind/brain, so how are they going to write a computer program that makes a computer/machine “understand”. When a human tells a story, there are emotions, gestures, etc. that are used, to convey and understand the story. Even animals show understanding. For example, a dog is happy to see its owner when he arrives, but may not show the same feelings for a stranger (except for the case where the dog is really friendly). The dog “understands” who its owner is and who it isn’t and displays emotion, based on this understanding. I know I’m not a dog so I can really speak for them, but from my “understanding” and observations of dogs. This is what I believe and I guess arguments could be different depending on your definition of understanding.  The following definitions were found on dictionary.com and I do not think that any one of these, but  a combination of all of them, make up an accurate definition of understanding.  Seeing as computers do not have emotions, they cannot understand.

un·der·stand
[uhn-der-stand] 
un·der·stood, un·der·stand·ing.
 
verb (used with object)
1. to perceive the meaning of; grasp the idea of; comprehend: to understand Spanish; I didn't understand your question.
2. to be thoroughly familiar with; apprehend clearly the character, nature, or subtleties of: to understand a trade.
3. to assign a meaning to; interpret: He understood her suggestion as a complaint.
4. to grasp the significance, implications, or importance of: He does not understand responsibility.
5. to regard as firmly communicated; take as agreed or settled: I understand that you will repay this loan in 30 days.
6. to learn or hear: I understand that you are going out of town.
7. to accept as true; believe: I understand that you are trying to be truthful, but you are wrong.
8. to construe in a particular way: You are to understand the phrase literally.
9. to supply mentally (something that is not expressed).

verb (used without object)
10. to perceive what is meant; grasp the information conveyed: She told them about it in simple words, hoping they would understand.
11. to accept tolerantly or sympathetically: If you can't do it, I'll understand.
12. to have knowledge or background, as on a particular subject: He understands about boats.
13. to have a systematic interpretation or rationale, as in a field or area of knowledge: He can repeat every rule in the book, but he just doesn't understand.


This was a really interesting topic to talk about and see what others opinions of it are. This blog post represents my opinions. If you disagree with me, that’s fine, you are entitled to your own opinion, too.

Monday, September 10, 2012

Design of Everyday Things - Ch. 1

Chapter 1: The Psychopathology of Everyday Things

I hate reading, but so far the book is somewhat interesting. I never really thought about why I decide to push/pull a door open. After reading this chapter, I realized how placing a handle/bar closer to one side of a door can subconsciously influence a person to push/pull in that direction. At my house, there are couple of clusters of switches in the kitchen, living room, and hallway, and although I have been living there for almost 6 years, I still have to push every switch to get the right set of lights to turn on. It gets frustrating, having to cycle through several switches to turn the right lights on. I have also realized that the design of a product is not simply thrown together. It is a careful, time-consuming process that is necessary to make sure an object has the ability to be used in the right way. I will probably trying to spot all kinds of bad designs, after reading this chapter.

Saturday, September 8, 2012

Assignment 1 - Paper Reading #6


Intro:
  • Paper reading #6: Handheld Augmented Reality Indoor Navigation with Activity-Based Instructions
Reference information:
  • Alessandro Mulloni, Hartmut Seichter, and Dieter Schmalstieg. 2011. Handheld augmented reality indoor navigation with activity-based instructions. In Proceedings of the 13th International Conference on Human Computer Interaction with Mobile Devices and Services (MobileHCI '11). ACM, New York, NY, USA, 211-220. DOI=10.1145/2037373.2037406 http://doi.acm.org.lib-ezproxy.tamu.edu:2048/10.1145/2037373.2037406
Author Bios: 
  • Alessandro Mulloni
    • Born in Udine, Italy, in 1981. 
    • Completed bachelors in computer science at the University of Milan
    • Received masters in computer science at the University of Udine
    • Studies focused on 3D real-time graphics and on Human-Computer Interaction
    • Worked as a researcher at the ITIA-CNR in Milan, inside the HCILab in Udine, and in the Handheld Augmented Reality group at the Graz University of Technology
    • Since 2008, working on PhD at the Christian Doppler Laboratory for Handheld Augmented Reality
     
  • Hartmut Seichter
    • 2003 - 2007, The University of Hong Kong, Architecture, PhD
    • 1996 - 2002, Bauhaus Universität Weimar, Architektur, Diplom Ingenieur
      

  •  Dieter Schmalstieg
    • Professor and head of the Institute for Computer Graphics and Vision at Graz University of Technology (TUG), Austria
    • Current research interests are augmented reality, virtual reality, real-time graphics, 3D user interfaces, and visualization
    • Received Dipl.-Ing. (1993), Dr. techn. (1997) and Habilitation (2001) degrees from Vienna University of Technology
    • Author and co-author of over 200 peer-reviewed scientific publications
    • Associate editor of IEEE Transactions on Visualization and Computer Graphics
    • Member of the editorial advisory board of computers & graphics and of Springer Virtual Reality
    • Member of the steering committee of the IEEE International Symposium on Mixed and Augmented Reality
    • Chair of the EUROGRAPHICS working group on Virtual Environments (1999-2010)

Summary:

A new design of an augmented reality interface, that supports indoor navigation, is used by combining activity-based instructions and sparse localization points to help with navigation in a building. Info points are used as checkpoints and provide AR information to the user in both text and WIM (world-in-miniature) maps. The user uses the instructions given by the information displayed on a handheld device to perform activities and the interface adapts the visualization, by changing the density/quality of the information shown. Users were tested with and without info points, to present the results of the effectiveness of the presence of info points and how they affect a users' performance on indoor navigation tasks. The results of this study show that info points act as confirmation points and provide an overview of the task, resulting in an improvement in task performance.



Related work not referenced in the paper:

1. Indoor navigation with mixed reality world-in-miniature views and sparse localization on mobile devices- Presents the design of an interface that provides continuous navigational support for indoor scenarios where localization is only available at sparse, discrete locations (info points) and uses an interface with turn-by-turn instructions and WIM maps. Extremely similar to the current paper.

2. Location based applications for mobile augmented reality - Talks about implementing indoor location based applications for a mobile AR system. Relevant to current paper because it uses a mobile device and augmented reality for indoor location based applications.

3. An experimental virtual museum based on augmented reality and navigation - Similar to current work because it describes an experiment using augmented reality to navigate a virtual museum.

4. Transformative reality: Augmented reality for visual prostheses - Uses a technique similar to AR called transformative reality where transformations are performed in real time to render low resolution vision for those who have visual handicaps. Also talks about indoor navigation. Cool idea for visually impaired.

5. Trends in augmented reality tracking, interaction and display: A review of ten years of ISMAR - Reviews the 10 year development of augmented reality at a ISMAR conference, with a focus on tracking, interaction, and display research. Looks into the future of AR and helps researchers decide which topics should be explored.

6. First Person Indoor/Outdoor Augmented Reality Application: ARQuake - This paper talks about a 1st person indoor/outdoor augmented reality application, called ARQuake, that is converted from the desktop game Quake. Relevant in the sense of augmented reality.

7. Outdoor augmented reality gaming on five dollars a day - Discusses hardware options available for a public AR gaming system and the reasons behind selecting appropriate sensors, head mounted displays, power sources, and controllers. Relevant to current work through augmented reality.

8. Outdoor See-Through Vision Utilizing Surveillance Cameras - Presents a novel outdoor mixed-reality system designed for those who carry a camera attached device outdoors, where several surveillance cameras are embedded. Relevant because of mixed-reality to be used as surveillance outdoors.

9. Interoperable augmented web browsing for exploring virtual media in real space - This paper focuses on the media for interoperable AR systems, assuming that these devices will become more wide spread and many people will share individual news in real world space. Discusses the use of augmented reality.

10. Building Virtual and Augmented Reality museum exhibitions - This paper talks about a system that allows museums to build/manage, both, virtual and augmented reality exhibits, based on 3D model representations of artifacts. Relevant because of the use of AR to create visualizations.

All of these works relate to the current work because they either use augmented reality in a way or have to do with the study of navigation techniques.


Evaluation:

The authors evaluated the work quantitatively, qualitatively, objectively, and subjectively. They measure the number of steps each participant took using a step counter and the amount of time it took to complete a task using a stop watch. The number of errors, both soft (user able to recover from mistake) and hard (not able to recover), were counted by one of the authors, too. The quantitative and subjective results are shown in the table below. The participants were also asked to fill out a NASA TLX questionnaire to evaluate the study based on the mental demand, physical demand, temporal demand, performance, effort, and frustration of the user using the system (with/without infor points. The qualitative/subjective results of the questionnaire are shown in the picture after the table. The AR navigation system was evaluated as a system. The study was conducted with/without info points and tasks that required all types of navigation instructions. The tasks were fairly similar in length and difficulty. This experiment tested every part of the AR navigation system.

                  Step difference | Time (seconds) | Soft | Hard | Total
NoIP                 29.75        |           135         |  19   |     3   |   22
IP                     -2.25          |          142          |   9   |     1   |   10 


Table 1. Task performance per condition: median difference
in step count from a pre-recorded ground truth, median task
completion time, and total number of navigation errors.






The results show that info points are extremely useful/helpful. Users felt that the info points reduced their workload and info points also reduced the amount of steps taken and mistakes made. Info points may have taken longer because time is spent at each of them browsing the displayed information.

Discussion: 

This paper presents good ideas and is novel because it is one of the first studies to implement augmented reality and activity-based instructions to help with navigation indoors. I believe work was evaluated appropriately because it tested performance with and without info points and it measured the study quantitatively and objectively, and qualitatively and subjectively. This idea could be really useful because it could help those who are in unfamiliar environments find their way. The info points are a really good idea, too, because humans like to confirm that what they are doing is right or if they are on the right track. Sparse localization is good because resources do not have to be wasted to set up a continuous infrastructure throughout a building. Overall, I liked the study and would like to see this implemented, especially in large buildings where their are no distinguishing landmarks within the building. Augmented reality was already a cool idea and using it to find objects would make it kind of fun.