Your SlideShare is downloading. ×
Blue eyes technology..document (1)
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×

Saving this for later?

Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime - even offline.

Text the download link to your phone

Standard text messaging rates apply

Blue eyes technology..document (1)

16,406
views

Published on


0 Comments
8 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total Views
16,406
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
1,259
Comments
0
Likes
8
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. ABSTRACTIs it possible to create a computer which can interact with us as we interact eachother? For example imagine in a fine morning you walk on to your computer roomand switch on your computer, and then it tells you “Hey friend, good morning youseem to be a bad mood today. And then it opens your mailbox and shows you some ofthe mails and tries to cheer you. It seems to be a fiction, but it will be the life lead by“BLUE EYES” in the very near future. The basic idea behind this technology is togive the computer the human power. We all have some perceptual abilities. That is wecan understand each other feelings. For example we can understand ones emotionalstate by analyzing his facial expression. If we add these perceptual abilities of humanto computers would enable computers to work together with human beings as intimatepartners. The “BLUE EYES” technology aims at creating computational machinesthat have perceptual and sensory ability like those of human beings. How can wemake computers "see" and "feel"? Blue Eyes uses sensing technology to identify ausers actions and to extract key information. This information is then analyzed todetermine the users physical, emotional, or informational state, which in turn can beused to help make the user more productive by performing expected actions or byproviding expected information. For example, in future a Blue Eyes-enabledtelevision could become active when the user makes eye contact, at which point theuser could then tell the television to "turn on". This paper is about the hardware,software, benefits and interconnection of various parts involved in the “blue eye”technology 1
  • 2. 1
  • 3. INDEX PAGE NO.CHAPTER-1INTRODUCTION TO BLUE EYES TECHNOLOGY 3-5CHAPTER-2AFFECTIVE COMPUTING 6-10CHAPTER-3MANUAL AND GAZE INPUT CASCADE (MAGIC) POINTING 11-18CHAPTER-4EYE TRACKER 19-21CHAPTER-5ARTIFICIAL INTELLIGENT SPEECH RECOGNITION 22-24CHAPTER-6APPLICATION OF BLUE EYES TECHNOLOGY 25-26CHAPTER-7ADVANTAGES OF BLUE EYES TECHNOLOGY 27CONCLUSION 28REFERENCES 29 2
  • 4. CHAPTER 1 INTRODUCTION TO BLUE EYES TECHNOLOGY Imagine yourself in a world where humans interact with computers.You are sitting in front of your personal computer that can listen, talk, or evenscream aloud. It has the ability to gather information about you and interact with youthrough special techniques like facial recognition, speech recognition, etc. It caneven understand your emotions at the touch of the mouse. It verifies your identity,feels your presents, and starts interacting with you. You ask the computer to dial to your friend at his office. It realizes theurgency of the situation through the mouse, dials your friend at his office, andestablishes a connection. The BLUE EYES technology aims at creating computationalmachines that have perceptual and sensory ability like those of human beings.Employing most modern video cameras and microphones to identifies the user’sactions through the use of imparted sensory abilities. The machine can understandwhat a user wants, where he is looking at, and even realize his physical or emotionalstates. The U.S. computer giant, IBM has been conducting research on theBlue Eyes technology at its Almaden Research Center (ARC) in San Jose, Calif.,since 1997. The ARC is IBMs main laboratory for basic research. The primaryobjective of the research is to give a computer the ability of the human being to assessa situation by using the senses of sight, hearing and touch. Animal survival dependson highly developed sensory abilities. Likewise, human cognition depends on highlydeveloped abilities to perceive, integrate, and interpret visual, auditory, and touchinformation. Without a doubt, computers would be much more powerful if they hadeven a small fraction of the perceptual ability of animals or humans. Adding suchperceptual abilities to computers would enable computers and humans to worktogether more as partners. Toward this end, the Blue Eyes project aims at creatingcomputational devices with the the sort of perceptual abilities that people take forgranted. Thus Blue eyes are the technology to make computers sense and understandhuman behavior and feelings and react in the proper ways. 3
  • 5. AIMS1) To design smarter devices2) To create devices with emotional intelligence3) To create computational devices with perceptual abilities The idea of giving computers personality or, more accurately, emotionalintelligence" may seem creepy, but the technologists say such machines would offerimportant advantages. De-spite their lightning speed and awesome powers of computation,todays PCs are essentially deaf, dumb, and blind. They cant see you, they cant hearyou, and they certainly dont care a whit how you feel. Every computer user knowsthe frustration of nonsensical error messages, buggy software, and abrupt systemcrashes. We might berate the computer as if it was an unruly child, but, of course, themachine cant respond. "Its ironic that people feel like dummies in front of theircomputers, when in fact the computer is the dummy," says Rosalind Picard, acomputer science professor at the MIT Media Lab in Cambridge. A computer endowed with emotional intelligence, on the other hand,could recognize when its operator is feeling angry or frustrated and try to respond inan appropriate fashion. Such a computer might slow down or replay a tutorialprogram for a confused student, or recognize when a designer is burned out andsuggest he take a break. It could even play a recording of Beethovens "MoonlightSonata" if it sensed anxiety or serve up a rousing Springsteen anthem if it detectedlethargy. The possible applications of "emotion technology" extend far beyond thedesktop. A car equipped with an affective computing system could recognizewhen a driver is feeling drowsy and ad-vise her to pull over, or it might sense when astressed-out motorist is about to explode and warn him to slow down and cool off. Human cognition depends primarily on the ability to perceive,interpret, and integrate audio-visuals and sensoring information. Adding extraordinaryperceptual abilities to computers would enable computers to work together withhuman beings as intimate partners.Researchers are attempting to add more capabilitiesto computers that will allow them to interact like humans, recognize human presents,talk, listen, or even guess their feelings. 4
  • 6. TRACKS USED Our emotional changes are mostly reflected in our heart pulserate, breathing rate, facial expressions, eye movements, voice etc. Hence these are theParameters on which blue technology is being developed. Making computers see and feel Blue Eyes uses sensing technology toidentify a users actions and to extract key information. This information is thenanalyzed to determine the users physical, emotional, or informational state, which inturn can be used to help make the user more productive by performing expectedactions or by providing expected information. Beyond making computers more researchers say there is anothercompelling reason for giving machine semotional intelligence. Contrary to thecommon wisdom that emotions contribute to irrational behavior, studies have shownthat feelings actually play a vital role in logical thought and decision- making.Emotionally impaired people often find it difficult to make decisions because they failto recognize the subtle clues and signals--does this make me feel happy or sad,excited or bored? That help direct healthy thought processes. It stands to reason,therefore, that computers that can emulate human emotions are more likely to behaverationally, in a manner we can understand. Emotions are like the weather. We onlypay attention to them when there is a sudden outburst, like a tornado, but in fact theyare constantly operating in the background, helping to monitor and guide ourday-to-day activities. Picard, who is also the author of the groundbreaking book AffectiveComputing, argues that computers should operate under the same principle."Theyhave tremendous mathematical abilities, but when it comes to interacting with people,they are autistic," she says. "If we want computers to be genuinely intelligent andinteract naturally with us, we must give them the ability to recognize, understand, andeven to behave and express emotions." Imagine the benefit of a computer that couldremember that a particular Internet search had resulted in a frustrating and futileexploration of cyberspace. Next time, it might modify its investigation to improve thechances of success when a similar request is made. 5
  • 7. CHAPTER 2 AFFECTIVE COMPUTING The process of making emotional computers with sensing abilitiesis known as affective computing. The steps used in this are:1)Giving sensing abilities2)Detecting human emotions3)Respond properly The first step, researchers say, is to give ma-chins the equivalent ofthe eyes, ears, and other sensory organs that humans use to recognize and expressemotion. To that end, computer scientists are exploring a variety of mechanismsincluding voice-recognition software that can discern not only what is being said butthe tone in which it is said; cameras that can track subtle facial expressions, eyemovements, and hand gestures; and biometric sensors that can measure bodytemperature, blood pressure, muscle tension, and other physiological signalsassociated with emotion. In the second step, the computers have to detect even the minorvariations of our moods. For e.g. person may hit the keyboard very fast either in thehappy mood or in the angry mood. In the third step the computers have to react in accordance MJ withthe emotional states. Various methods of accomplishing affective computing are:1) AFFECT DETECTION.2) MAGIC POINTING.3) SUITOR.4) EMOTIONAL MOUSE. 6
  • 8. 1) AFFECT DETECTION This is the method of detecting our emotional states from theexpressions on our face. Algorithms amenable to real time implementation that extractinformation from facial expressions and head gestures are being explored. Most of theinformation is extracted from the position of the eye rows and the corners of themouth.2) MAGIC POINTING MAGIC stands for Manual Acquisition with Gaze TrackingTechnology. a computer with this technology could move the cursor by following thedirection of the users eyes. This type of technology will enable the computer toautomatically transmit information related to the screen that the user is gazing at.Also, it will enable the computer to determine, from the users expression, if he or sheunderstood the information on the screen, before automatically deciding to proceed tothe next program. The user pointing is still done by the hand, but the cursor alwaysappears at the right position as if by MAGIC. By varying input technology and eyetracking, we get MAGIC pointing.3) SUITOR SUITOR stands for Simple User Interface Tracker. It implementsthe method for putting computational devices in touch with their users changingmoods. By watching what we page the user is currently browsing, the SUITOR canfind additional information on that topic. The key is that the user simply interacts withthe computer as usual and the computer infers user interest based on what it sees theuser do.4) EMOTION MOUSE This is the mouse embedded with sensors that can sense thephysiological attributes such as temperature, Body pressure, pulse rate, and touchingstyle, etc. The computer can determine the user’s emotional states by a single touch.IBM is still Performing research on this mouse and will be available in the marketwithin the next two or three years. The expected accuracy is 75%. One goal of human computer interaction (HCI) is to make anadaptive, smart computer system. This type of project could possibly include gesture 7
  • 9. recognition, facial recognition, eye tracking, speech recognition, etc. Another non-invasive way to obtain information about a person is through touch. People use theircomputers to obtain, store and manipulate data using. In order to start creating smart computers, the computer must startgaining information about the user. Our proposed method for gaining user informationthrough touch is via a computer input device, the mouse. From the physiological dataobtained from the user, an emotional state may be determined which would then berelated to the task the user is currently doing on the computer. Over a period of time, auser model will be built in order to gain a sense of the users personality. The scope of the project is to have the computer adapt to the user inorder to create a better working environment where the user is more productive. Thefirst steps towards realizing this goal are described here.2.1. EMOTION AND COMPUTING Rosalind Picard (1997) describes why emotions are important to thecomputing community. There are two aspects of affective computing: giving thecomputer the ability to detect emotions and giving the computer the ability to expressemotions. Not only are emotions crucial for rational decision making but emotiondetection is an important step to an adaptive computer system. An adaptive, smartcomputer system has been driving our efforts to detect a person’s emotional state. By matching a person’s emotional state and the context of theexpressed emotion, over a period of time the person’s personality is being exhibited.Therefore, by giving the computer a longitudinal understanding of the emotional stateof its user, the computer could adapt a working style which fits with its user’spersonality. The result of this collaboration could increase productivity for the user.One way of gaining information from a user non-intrusively is by video. Camerashave been used to detect a person’s emotional state. We have explored gaininginformation through touch. One obvious place to put sensors is on the mouse. 8
  • 10. Figure 2.1.physiological emotions for emotional mouse 9
  • 11. 2.2. THEORY Based on Paul Elman’s facial expression work, we see a correlationbetween a person’s emotional state and a person’s physiological measurements.Selected works from Elman and others on measuring facial behaviors describeElman’s Facial Action Coding System (Elman and Rosenberg, 1997). One of his experiments involved participants attached to devices torecord certain measurements including pulse, galvanic skin response (GSR),temperature, somatic movement and blood pressure. He then recorded themeasurements as the participants were instructed to mimic facial expressions whichcorresponded to the six basic emotions. He defined the six basic emotions as anger,fear, sadness, disgust, joy and surprise. From this work, Dryer (1993) determined howphysiological measures could be used to distinguish various emotional states. Themeasures taken were GSR, heart rate, skin temperature and general somatic activity(GSA). These data were then subject to two analyses. For the first analysis, amultidimensional scaling. (MDS) procedure was used to determine the dimensionalityof the data.2.3. RESULT The data for each subject consisted of scores for four physiologicalassessments [GSA, GSR, pulse, and skin temperature, for each of the six emotions(anger, disgust, fear, happiness, sadness, and surprise)] across the five minute baselineand test sessions. GSA data was sampled 80 times per second, GSR and temperaturewere reported approximately 3-4times per second and pulse was recorded as a beatwas detected, approximately 1 time per second. To account for individual variance inphysiology, we calculated the difference between the baseline and test scores. Scoresthat differed by more than one and a half standard deviations from the mean weretreated as missing. By this criterion, twelve score were removed from the analysis.The results show the theory behind the Emotion mouse work is fundamentally sound. 10
  • 12. CHAPTER 3 MANUAL AND GAZE INPUT CASCADED (MAGIC) POINTING This work explores a new direction in utilizing eye gaze for computerinput. Gaze tracking has long been considered as an alternative or potentially superiorpointing method for computer input. We believe that many fundamental limitationsexist with traditional gaze pointing. In particular, it is unnatural to overload aperceptual channel such as vision with a motor control task. We therefore propose analternative approach, dubbed MAGIC (Manual and Gaze Input Cascaded) pointing.With such an approach, pointing appears to the user to be a manual task, used for finemanipulation and selection. However, a large portion of the cursor movement iseliminated by warping the cursor to the eye gaze area, which encompasses the target. Two specific MAGIC pointing techniques, one conservative andone liberal, were designed, analyzed, and implemented with an eye tracker wedeveloped. They were then tested in a pilot study. This early stage exploration showedthat the MAGIC pointing techniques might offer many advantages, including reducedphysical effort and fatigue as compared to traditional manual pointing, greateraccuracy and naturalness than traditional gaze pointing, and possibly faster speed thanmanual pointing. In our view, there are two fundamental shortcomings to theexisting gaze pointing techniques, regardless of the maturity of eye trackingtechnology First, given the one-degree size of the fovea and the subconsciousjittery motions that the eyes constantly produce, eye gaze is not precise enough tooperate UI widgets such as scrollbars, hyperlinks, and slider handles Second, andperhaps more importantly, the eye, as one of our primary perceptual devices, has notevolved to be a control organ. Sometimes its movements are voluntarily controlledwhile at other times it is driven by external events. With the target selection by dwelltime method, considered more natural than selection by blinking [7], one has to beconscious of where one looks and how long one looks at an object. If one does not 11
  • 13. look at a target continuously for a set threshold (e.g., 200ms), the target will not besuccessfully selected. Once the cursor position had been redefined, the user would need toonly make a small movement to, and click on, the target with a regular manual inputdevice. We have designed two MAGIC pointing techniques, one liberal and the otherconservative in terms of target identification and cursor placement. The liberal MAGIC pointing technique: cursor is placed in thevicinity of a target that the user fixates on. Actuate input device, observe the cursorposition and decide in which direction to steer the cursor. The cost to this method isthe increased manual movement amplitude. The conservative MAGIC pointing technique with "intelligentoffset" To initiate a pointing trial, there are two strategies available to the user. One isto follow "virtual inertia:" move from tie cursors current position towards the newtarget the user is looking at. This is likely the strategy the user will employ, due to theway the user interacts with todays interface. The alternative strategy, which may bemore advantageous but takes time to learn, is to ignore the previous cursor positionand make a motion which is most convenient and least effortful to the user for a giveninput device. The goal of the conservative MAGIC pointing method is thefollowing. Once the user looks at a target and moves the input device, the cursor willappear "out of the blue" in motion towards the target, on the side of the targetopposite to the initial actuation vector. In comparison to the liberal approach, thisconservative approach has both pros and cons. While with this technique the cursorwould never be over-active and jump to a place the user does not intend to acquire, itmay require more hand-eye coordination effort. Both the liberal and the conservativeMAGIC pointing techniques offer the following potential advantages:1. Reduction of manual stress and fatigue, since the cross screen long-distance cursormovement is eliminated from manual control.2. Practical accuracy level. In comparison to traditional pure gaze pointing whoseaccuracy is fundamentally limited by the nature of eye movement, the MAGIC 12
  • 14. pointing techniques let the hand complete the pointing task, so they can be as accurateas any other manual input techniques.3. A more natural mental model for the user. The user does not have to be aware ofthe role of the eye gaze. To the user, pointing continues to be a manual task, with acursor conveniently appearing where it needs to be.4. Speed. Since the need for large magnitude pointing operations is less than with puremanual cursor control, it is possible that MAGIC pointing will be faster than puremanual pointing.5. Improved subjective speed and ease-of-use. Since the manual pointing amplitude issmaller, the user may perceive the MAGIC pointing system to operate faster and morepleasantly than pure manual control, even if it operates at the same speed or moreslowly. The fourth point wants further discussion. According to the wellaccepted Fits Law, manual pointing time is logarithmically proportional to the A/Wratio, where A is the movement distance and W is the target size. In other words,targets which are smaller or farther away take longer to acquire. For MAGIC pointing, since the target size remains the same but thecursor movement distance is shortened, the pointing time can hence be reduced. It isless clear if eye gaze control follows Fits Law. In Ware and Michelin’s study,selection time was shown to be logarithmically proportional to target distance, therebyconforming to Fits Law. To the contrary, Silber and Jacob [9] found that trialcompletion time with eye tracking input increases little with distance, thereforedefying Fits Law. In addition to problems with todays eye tracking systems, such asdelay, error, and inconvenience, there may also be many potential human factordisadvantages to the MAGIC pointing techniques we have proposed, including thefollowing:1. With the more liberal MAGIC pointing technique, the cursor warping can beoveractive at times, since the cursor moves to the new gaze location whenever the eyegaze moves more than a set distance (e.g., 120 pixels) away from the cursor. Thiscould be particularly distracting when the user is trying to read. It is possible to 13
  • 15. introduce additional constraint according to the context. For example, when the userseye appears to follow a text reading pattern, MAGIC pointing can be automaticallysuppressed.2. With the more conservative MAGIC pointing technique, the uncertainty of theexact location at which the cursor might appear may force the user, especially anovice, to adopt a cumbersome strategy: take a touch (use the manual input device toactivate the cursor), wait (for the cursor to appear), and move (the cursor to the targetmanually). Such a strategy may prolong the target acquisition time. The user mayhave to learn a novel hand-eye coordination pattern to be efficient with this technique.Gaze position reported by eye tracker Eye tracking boundary with 95% confidenceTrue target will be within the circle with 95% probability The cursor is warped to theboundary of the gaze area, along the initial actuation vector Previous cursor position,far from target Initial manual actuation vector3. With pure manual pointing techniques, the user, knowing the current cursorlocation, could conceivably perform his motor acts in parallel to visual search. Motoraction may start as soon as the users gaze settles on a target. With MAGIC pointingtechniques, the motor action computation (decision) cannot start until the cursorappears. This may negate the time saving gained from the MAGIC pointingtechniques reduction of movement amplitude. Clearly, experimental (implementationand empirical) work is needed to validate, refine, or invent alternative MAGICpointing techniques. 14
  • 16. 3.1.1. ADVANTAGES OF LIBERAL CONSERVATIVE APPROACH 1. Reduction of manual stress and fatigue 2. Practical accuracy level 3. A more natural mental model for the user 4. Faster than pure manual pointing 5. Improved subjective speed and ease of use3.1.2. DISDVANTAGES IN LIBERAL CONSERVATIVE APPROACH 1. Liberal approach is distracting when the user is trying to read 2. The motor action computation cannot start until the cursor appears 3. In conservative approach, uncertainty of the exact location prolong the target acquisition time3.2. IMPLEMENTING MAGIC POINTING We programmed the two MAGIC pointing techniques on a WindowsNT system. The techniques work independently from the applications. The MAGICpointing program takes data from both the manual input device (of any type, such as amouse) and the eye tracking system running either on the same machine or on anothermachine connected via serial port. Raw data from an eye tracker cannot be directlyused for gaze-based interaction, due to noise from image processing, eye movementjitters, and samples taken during Saccade (ballistic eye movement) periods. The goal of filter design in general is to make the best compromisebetween preserving signal bandwidth and eliminating unwanted noise. In the case ofeye tracking, as Jacob argued, eye information relevant to interaction lies in thefixations3.3. EXPERIMENT Empirical studies are relatively rare in eye tracking-based interactionresearch, although they are particularly needed in this field. Human behavior andprocesses at the perceptual motor level often do not conform to conscious-levelreasoning. One usually cannot correctly describe how to make a turn on a bicycle.Hypotheses on novel interaction techniques can only be validated by empirical data. 15
  • 17. However, it is also particularly difficult to conduct empirical research on gaze-basedinteraction techniques, due to the complexity of eye movement and the lack ofreliability in eye tracking equipment. Satisfactory results only come when "everythingis going right." When results are not as expected, It is difficult to find the true reasonamong many possible reasons: Is it because a subjects particular eye property fooledthe eye tracker Was there a calibration error Or random noise in the imaging systemOr is the hypothesis in fact invalid We are still at a very early stage of exploring theMAGIC pointing techniques. More refined or even very different techniques may bedesigned in the future. We are by no means ready to conduct the definitive empiricalstudies on MAGIC pointing. However, we also feel that it is important to subject ourwork to empirical evaluations early so that quantitative observations can be made andfed back to the iterative design-evaluation-design cycle. We therefore decided toconduct a small-scale pilot study to take an initial peek at the use of MAGIC pointing,however unrefined3.3.1. EXPERIMENTAL DESIGN The two MAGIC pointing techniques described earlier were put to testusing a set of parameters such as the filters temporal and spatial thresholds, theminimum cursor warping distance, and the amount of "intelligent bias" (subjectivelyselected by the authors without extensive user testing). Ultimately the MAGICpointing techniques should be evaluated with an array of manual input devices,against both pure manual and pure gaze-operated pointing methods.Since this is an early pilot study, we decided to limit ourselves to one manual inputdevice. A standard mouse was first considered to be the manual input device in theexperiment. However, it was soon realized not to be the most suitable device forMAGIC pointing, especially when a user decides to use the push-upwards strategywith the intelligent offset. Because in such a case the user always moves in onedirection, the mouse tends to be moved off the pad, forcing the user adjust the mouseposition, often during a pointing trial. We hence decided to use a miniature isometricpointing stick (IBM Track Point IV commercially used in the IBM ThinkPad 600 and770 series notebook computers). Another device suitable for MAGIC pointing is atouchpad: the user can choose one convenient gesture and to take advantage of theintelligent offset. The experimental task was essentially a Fits pointing task. Subjectswere asked to point and click at targets appearing in random order. If the subject 16
  • 18. clicked off-target, a miss was logged but the trial continued until a target was clicked.An extra trial was added to make up for the missed trial. Only trials with no misseswere collected for time performance analyses. Subjects wereasked to complete the task as quickly as possible and as accurately as possible. Toserve as a motivator, a $20 cash prize was set for the subject with the shortest meansession completion time with any technique. The task was presented on a 20 inch CRT color monitor, with a 15 by 11inch viewable area set at resolution of 1280 by 1024 pixels. Subjects sat from thescreen at a distance of 25 inches. The following factors were manipulated in theexperiments: two target sizes: 20 pixels (0.23 in or 0.53 degree of viewing angle at 25in distance) and 60 pixels in diameter (0.7 in, 1.61degree)three target distances: 200pixels (2.34 in, 5.37 degree), 500 pixels (5.85 in, 13.37 degree), and 800 pixels (9.38in, 21.24degree)three pointing directions: horizontal, vertical and diagonal. A within-subject design was used. Each subject performed the task with all three techniques:(1) Standard, pure manual pointing with no gaze tracking (No Gaze);(2) The conservative MAGIC pointing method with intelligent offset (Gaze);(3) The liberal MAGIC pointing method (Gaze2). Nine subjects, seven male and two female, completed theexperiment. The order of techniques was balanced by a Latin square pattern. Sevensubjects were experienced Track Point users, while two had little or no experience.With each technique, a 36-trial practice session was first given, during which subjectswere encouraged to explore and to find the most suitable strategies (aggressive,gentle, etc.). The practice session was followed by two data collection sessions.Although our eye tracking system allows head motion, at least for those users who donot wear glasses, we decided to use a chin rest to minimize instrumental error.3.3.2. EXPERIMENTAL RESULTSSession Given the pilot nature and the small scale of the experiment, we expectedthe statistical power of the results to be on the weaker side. In other words, while thesignificant effects revealed are important, suggestive trends that are statistically non-significant are still worth noting for future research. 17
  • 19. Session2 Mean completion time (sec) vs. experiment sessionThe total average completion time was 1.4 seconds with the standard manual controltechnique 1.52 seconds with the conservative MAGIC pointing technique (Gaze), and1.33 seconds with the liberal MAGIC pointing technique (Gaze2). Note that the GazeTechnique had the greatest improvement from the first to the second experimentsession, suggesting the possibility of matching the performance of the other twotechniques with further practice. As expected, target size significantly influenced pointing time: f(1,8) =178, p < 0.001. This was true for both the manual and the two MAGIC pointingtechniques. Pointing amplitude also significantly affected completion time: F(2, 8) =97.5, p < 0.001. However, the amount of influence varied with the technique used, asindicated by the significant interaction between technique and amplitude: F(4, 32) =7.5,p<0.001. As pointing amplitude increased from 200 pixels to 500 pixels and thento 800 pixels, subjects completion time with the No Gaze condition increased in anon-linear, logarithmic-like pace as Fits Law predicts. This is less true with thetwo MAGIC pointing techniques, particularly the Gaze2 condition, which is definite:,not logarithmic. Nonetheless, completion time with the MAGIC pointing techniquesdid increase as target distance increased. This is intriguing because in MAGICpointing techniques, the manual control portion of the movement should be thedistance from the warped cursor position to the true target. Such distance depends oneye tracking system accuracy, which is unrelated to the previous cursor position. In short, while completion time and target distance with the MAGICpointing techniques did not completely follow Fits Law, they were not completelyindependent either. Indeed, when we lump target size and target distance according tothe Fits Law Index of Difficulty ID = log2(A/W+1) [15], we see a similarphenomenon. 18
  • 20. CHAPTER 4 EYE TRACKER Figure 4.1 The liberal MAGIC pointing technique: the curser is placed in the vicinity of the target that the user fixates onFigure 4.2.The conservative MAGIC pointing technique with “intelligent offset” 19
  • 21. Since the goal of this work is to explore MAGIC pointing as auser interface technique, we started out by purchasing a commercial eye tracker (ASLModel 5000) after a market survey. In comparison to the system reported in earlystudies this system is much more compact and reliable. However, we felt that it wasstill not robust enough for a variety of people with different eye characteristics, suchas pupil brightness and correction glasses. We hence chose to develop and use our own eye tracking system.Available commercial systems, such as those made by ISCAN Incorporated,LC Technologies, and Applied Science Laboratories(ASL), rely on a single lightsource that is positioned either off the camera axis in the case of the ISCANETL-400systems, or on-axis in the case of the LCT and the ASL E504 systems.Figure 4.3.Bright (left) and dark (right) pupil images resulting from on-axis andoff-axis illumination. The glints, or corneal reflections, from the on- and off-axis light sources can be easily identified as the bright points in the iris. Eye tracking data can be acquired simultaneously with MRIscanning using a system that illuminates the left eye of a subject with an infrared (IR)source, acquires a video image of that eye, locates the corneal reflection (CR) of theIR source, and in real time calculates/displays/records the gaze direction and pupildiameter. 20
  • 22. Figure 4.4.MRI scanning Once the pupil has been detected, the corneal reflection isdetermined from the dark pupil image. The reflection is then used to estimate theusers point of gaze in terms of the screen coordinates where the user is looking at. Aninitial calibration procedure, similar to that required by commercial eye trackers. 21
  • 23. CHAPTER 5 ARTIFICIAL INTELLIGENCE SPEECH RECOGNITION It is important to consider the environment in which the speechrecognition system has to work. The grammar used by the speaker and accepted bythe system, noise level, noise type, position of the microphone, and speed and mannerof the user’s speech are some factors that may affect the quality of speech recognition.When you dial the telephone number of a big company, you are likely to hear thesonorous voice of a cultured lady who responds to your call with great courtesysaying “Welcome to company X. Please give me the extension number you want”.You pronounce the extension number, your name, and the name of person you want tocontact. If the called person accepts the call, the connection is given quickly. This isartificial intelligence where an automatic call-handling system is used withoutemploying any telephone operator.. 5.1 THE TECHNOLOGY Artificial intelligence (Al) involves two basic ideas. First, itinvolves studying the thought processes of human beings. Second, it deals withrepresenting those processes via machines (like computers, robots, etc). Al is behaviorof a machine, which, if performed by a human being, would be called intelligent. Itmakes machines smarter and more useful, and is less expensive than naturalintelligence. Natural language processing (NLP) refers to artificial intelligencemethods of communicating with a computer in a natural language like English. Themain objective of a NLP program is to understand input and initiate action. The inputwords are scanned and matched against internally stored known words. Identificationof a key word causes some action to be taken. In this way, one can communicate withthe computer in ones language. No special commands or computer language arerequired. There is no need to enter programs in a special language for creatingsoftware. 22
  • 24. 5.2 SPEECH RECOGNITION The user speaks to the computer through a microphone, which,in used; a simple system may contain a minimum of three filters. The more thenumber of filters used, the higher the probability of accurate recognition. Presently,switched capacitor digital filters are used because these can be custom-built inintegrated circuit form. These are smaller and cheaper than active filters usingoperational amplifiers. The filter output is then fed to the ADC to translate the analoguesignal into digital word. The ADC samples the filter outputs many times a second.Each sample represents different amplitude of the signal .Evenly spaced vertical linesrepresent the amplitude of the audio filter output at the instant of sampling. Eachvalue is then converted to a binary number proportional to the amplitude of thesample. A central processor unit (CPU) controls the input circuits that are fed by the ADCS. A large RAM (random access memory) stores all thedigital values in a buffer area. This digital information, representing the spoken word,is now accessed by the CPU to process it further. The normal speech has a frequencyrange of 200 Hz to 7 kHz. Recognizing a telephone call is more difficult as it hasbandwidth limitation of 300Hz to 3.3 kHz. As explained earlier, the spoken words are processed by the filtersand ADCs. The binary representation of each of these words becomes a template orstandard, against which the future words are compared. These templates are stored inthe memory. Once the storing process is completed, the system can go into its activemode and is capable of identifying spoken words. As each word is spoken, it isconverted into binary equivalent and stored in RAM. The computer then startssearching and compares the binary input pattern with the templates, t is to be notedthat even if the same speaker talks the same text, there are always slight variations inamplitude or loudness of the signal, pitch, frequency difference, time gap, etc. Due tothis reason, there is never a perfect match between the template and binary inputword. The pattern matching process therefore uses statistical techniques and isdesigned to look for the best fit. The values of binary input words are subtracted from thecorresponding values in the templates. If both the values are same, the difference iszero and there is perfect match. If not, the subtraction produces some difference orerror. The smaller the error, the better the match. When the best match occurs, the 23
  • 25. word is identified and displayed on the screen or used in some other manner. Thesearch process takes a considerable amount of time, as the CPU has to make manycomparisons before recognition occurs. This necessitates use of very high-speedprocessors. A large RAM is also required as even though a spoken word may last onlya few hundred milliseconds, but the same is translated into many thousands of digitalwords. It is important to note that alignment of words and templates are to be matchedcorrectly in time, before computing the similarity score. This process, termed asdynamic time warping, recognizes that different speakers pronounce the same wordsat different speeds as well as elongate different parts of the same word. This isimportant for the speaker-independent recognizers. 24
  • 26. CHAPTER 6 APPLICATIONS OF BLUE EYE TECHNOLOGY One of the main benefits of speech recognition system is that it letsuser do other works simultaneously. The user can concentrate on observation andmanual operations, and still control the machinery by voice input commands. Anothermajor application of speech processing is in military operations. Voice control ofweapons is an example. With reliable speech recognition equipment, pilots can givecommands and information to the computers by simply speaking into theirmicrophones—they don’t have to use their hands for this purpose. Another goodexample is a radiologist scanning hundreds of X-rays, ultra sonograms, CT scans andsimultaneously dictating conclusions to a speech recognition system connected toword processors. The radiologist can focus his attention on the images rather thanwriting the text. Voice recognition could also be used on computers for making airlineand hotel reservations. A user requires simply stating his needs, to make reservation,cancel a reservation, or making enquiries about schedule.6.1. APPLICATIONS OF ARTIFICIAL SPEECH RECOGNITION 1. To control weapons by voice commands 2. Pilot give commands to computers by speaking into microphones 3. For making airline and hotel reservations 4. For making reservations, canceling reservations or making enquiries 5. Can be connected to word processors and instead of writing, simply dictate to them 25
  • 27. Some of the blue Eyes enabled devices are discussed below:1)POD: The first blue Eye enabled mass production device was POD, the carmanufactured y TOYOTA. It could keep the driver alert and active. It could tell the driver togo slow if he is driving too fast and it could pull over the driver when he feels drowsy. Also itcould hear the driver some sort of interesting music when he is getting bored.2) PONG: IBM released a robot designed for demonstrating the newtechnology. The Blue Eyes robot is equipped with a computer capable of analyzing apersons glances and other forms of expressions of feelings, before automaticallydetermining the next type of action. IBM has released a robot called PONG, which isequipped with the Blue Eyes technology. PONG is capable of perceiving the personstanding in front of it, smiles when the person calls his name, and expresses lonelinesswhen it loses sight of the person. 26
  • 28. CHAPTER 7 ADVANTAGES OF BLUE EYE TECHNOLOGYThe simple user interface tracker Computers would have been much more powerful, had theygained perceptual and sensory abilities of the living beings on the earth. What needsto be developed is an intimate relationship between the computer and the humans.And the Simple User Interest Tracker (SUITOR) is a revolutionary approach in thisdirection. By observing the Webpage at bedizen is browsing, the SUITOR can help byfetching more information at his desktop. By simply noticing where the user’s eyesfocus on the computer screen, the SUITOR can be more precise in determining histopic of interest. The Almaden cognitive scientist who invented SUITOR, "thesystem presents the latest stock price or business news stories that could affect IBM.If I read the headline off the ticker, it pops up the story in a browser window. If I startto read the story, it adds related stories to the ticker. Thats the whole idea of anattentive system—one that attends to what you are doing, typing, reading, so that itcan attend to your information needs.SUITOR:- 1. Help by fetching more information at desktop 2. Notice where the user’s eyes focus on the screen 3. Fills a scrolling ticker on a computer screen with information related to user’s task 27
  • 29. CONCLUSION: The nineties witnessed quantum leaps interface designing for improvedman machine interactions. The BLUE EYES technology ensures a convenient way ofsimplifying the life by providing more delicate and user friendly facilities incomputing devices. Now that we have proven the method, the next step is to improvethe hardware. Instead of using cumbersome modules to gather information about theuser, it will be better to use smaller and less intrusive units. The day is not far whenthis technology will push its way into your house hold, making you more lazy. It mayeven reach your hand held mobile device. Any way this is only a technologicalforecast. 28
  • 30. REFERENCES: 1). Levin J.L, An eye controlled computers. 2). Silbert.L and R.Jacob, The advantage of eye gazing interactions. 3). Richard A.Bolt, Eyes at interface. 4).Colin ware, Harutune H.Mikaelian, An evalution of eye tracker. 29