10-04-2017, 08:06 PM
Is that all that is left to be done?
Have we reached a state where we have almost invented all that we need?
Is our target only to keep crowding more and more transistors on to smaller and smaller I.Cs?
To write millions and billions of lines of code hoping to solve all of mankind's problems. Is there nothing else to do?
Perceptive Computing
Presented By,
DILEEP R.A
What is perception?
The process of acquiring knowledge through the senses.
Ablility to feel, see and sense.
What is a perceptive computer?
A computer with perceptual capabilities.
Abilities to perceive, integrate, and interpret visual, auditory and touch information.
Just like communicating with a human.
Why perception?
To make the human computer interaction more -
Natural
Easier
Richer
Computers would be much more powerful if they had even a small fraction of the perceptual ability of humans.
Perceptive Computing Projects
Project BlueEyes
(IBM Almaden Computer Science Research Lab)
Multimodal Interfaces
(The Human Technology Interface Lab, Washington University)
Features
Uses non-obtrusive technology
Extracts key information from the user s cues and gestures.
These cues are analysed to determine the user s physical, emotional or informational state.
Attentive environments environments that are user and context aware.
BlueEyes Projects
Emotion mouse
Suitor
MAGIC
Facial expression identification
Features
Simply by touching or using the mouse normally the computer will be able to determine the person s emotional state
The emotional state can then be related to the task the user is currently doing on the computer.
Over time, a user model can be built up and the computer can adapt to the user to create a better working environment.
Working
Working (contd)
The IR, GSA and GSR are inputs into a series of discriminate function analyses.
The result is correlated to an emotional state.
Hence, now an association has been established the user s physiological state and the associated state.
Facial Expression Identification
Assesses emotions by using the image of the person.
Simple states like anxious and happy can be assessed easily.
Complex states still being researched.
MAGIC (Manual Acquisition with Gaze Initiated Control)
Uses gaze tracking to perceive user s area of interest
The mouse cursor moves to the area of the screen that the user gazes at.
Still mouse button is used for selection since the accuracy of the eye is not good.
SUITOR (Simple User Interest Tracker)
Puts computational devices in touch with their user s changing information needs.
Suitor can infer what kind of information will be interesting to the user at a particular time and deliver it.
Intelligent data gathering and data mining.
Multimodal Interfaces
Development of software libraries for incorporating multimodal input into human computer interfaces.
Allow human-computer interaction with an intuitive mix of voice, gesture, speech, gaze and body motion.
Projects
Intelligent Conversational Avatar
HMRS (Hand Motion Recognition Sytem)
Intelligent Conversational Avatar
Aims
Develop an Expert System and Natural Language Parsing module to parse emotive expressions from textual input.
This information is used to set the graphical appearance of avatars.
Model
Aims
To develop a generic software package for hand motion recognition using hidden markov models.
Using this user interface designers will be able to build a multimodal input system.
Uses
Specifying shapes and sizes in CAD/CAM applications
Hand posture command inputs
Menu selections
Where?
Computers
Embedded devices
PDAs
Mobile phones
Human interface devices
Present
Voice recognition
dialing in mobile phones
commands for PCs
Data mining used by websites
Future
Computer training or educational programs
Interactive entertainment
Advertising
User based search engines
Future (contd)
More responsive user interfaces
Household appliances that we can talk to
Mood perceptible software
Possibilities are endless
Conclusion
Looking to the future, we can predict that computers will be equipped with varying degrees of perception.
The presented perception projects aim to implement those capabilities to make computers more human in sensing.
They will enhance the way a human is able to instruct a computer to do a task.