Gesture Analysis
Duration: Half day
Schedule:
Saturday, 20th March 2010, Afternoon
Fee:
Rs. 3,000
Participants: 10-25
Aim
Why Gestures, and why hand gestures?
Gestures are a natural modality for human-computer interaction
(HCI). Why gestures, and not other non-keyboard and mouse-based
interaction modality? Speech-based systems have a restriction of
of a good training set, and the efficiency of such systems
crucially depends on the training set. Haptic interfaces are
hands-on, and based on touch - which makes them somewhat limited
in their scope. Hand gestures are a natural mode of
communication in most cultures, and are often used as a language
supplement. Gestures are a natural mode of communication for
hearing- and speech-impaired individuals.
Content
In this tutorial, we examine gestures, and in particular,
hand
gestures based on visual inputs. Advances in the field of Computer
Vision and Image analysis have
enabled estimation of 3-D structure of objects with calibrated
multi-camera systems. However, since the 1990s, there has been a
trend towards practical computer vision applications: simple
systems without any cumbersome setup, using off-the-shelf items:
cameras, and phones, for instance. Another trend has been to do
away with a trained operator, for automatic/semi-automatic
computer vision systems. Calibration is often cumbersome, and
requires trained operators. There is a trend towards natural
modes of interaction with computers/cameras/other devices.
The challenges in visual analysis of hand gestures with an uncalibrated cameras include the fact that hands are highly articulated objects, of varying sizes and shapes. They change both their position and appearance across frames. We consider modelling and recognising spatio-temporal hand gestures - having same shapes following different trajectories, and different shapes following the same trajectories.
Mobile devices with a large number of sensors are becoming
ubiquitous now-a-days. Mobile phones for instance, are coming
with in-built sensors such as cameras and accelerometers. With
such mobile devices used in the hand, they often form a
convenient modality for hand gestures, which need to be captured
and automatically interpreted by mobile devices with sensors such
as cameras and accelerometers. The tutorial will also examine
gestures made with accelerometer-based sensors. Another
interesting class of gestures are those with pen-based inputs. Electronic
pens are also becoming a popular input sensor.
Stylus-based input sensors have been around for mobile phones and
laptops/notebooks. Electronic pens and tablets in combination
enable the use of gestures with pens being another input
modality. Electronic pens also come with camera-based inputs to
scan input on conventional paper. Having different input
modalities offers a rich testbed to experiment with different
types of gestures. This opens up a vast array of possibilities
for multimedia communication with different modalities: combining
asynchronous and synchronous modalities.
Who should attend?
This tutorial will be technology oriented - will have hands-on
experience as well as quantitative research from a pattern
recognition and image analysis point of view. The tutorial will
be useful both for technology-oriented people as well as those
interested in a user perspective using common input devices.
The tutorial welcomes people with backgrounds in Human-Computer
Interaction, Computer Science and Electrical and Electronic Engineering.
Short biographies of the tutors
Sumantra
Dutta Roy
Sumantra Dutta Roy is a B.E. (Computer Engineering) from D.I.T.,
Delhi, and completed his M.Tech and Ph.D. degrees at the
Department of Computer Science and Engineering, I.I.T. Delhi. He
started his career in teaching and research in the Department of
Electrical Engineering at I.I.T. Bombay, where he worked from
2001 to early 2007. Since then, he has joined I.I.T. Delhi in the
Department of Electrical Engineering as an Associate Professor.
He is a recepient of 2004 INAE Young Engineer Award (Indian
National Academy of Engineering), and the 2004 - 05 BOYSCAST
Fellowship of the Department of Science and Technology,
Government of India. His research interests are in Computer
Vision and Image Processing, Music Information Retrieval, and
Bioinformatics.
http://www.cse.iitd.ac.in/~sumantra
Santanu Chaudhury
Santanu Chaudhury did his B.Tech (1984) in Electronics and Electrical Communication
Engg and Ph.D (1989) in Computer
Science and Engg. from I.I.T, Kharagpur, India. Currently, he is
a professor in the department of Electrical Engg. at
I.I.T. Delhi. He was awarded INSA medal for Young Scientists, and
is a Fellow of the Indian National Academy of Engineers and the
National Academy of Sciences.
His research interests are in the areas of Computer Vision,
Artificial Intelligence and Multimedia Systems. He has published
more than 100 papers in international journals and conference
proceedings.
http://www.iitd.ac.in/deptt/ee/people/faculty/schaudhury.html
Back to Workshops & Tutorials
