Rules‎ > ‎

Goal

The goal of the challenge is to develop a system capable of recognizing hand gestures, signs, or signals, from video data. Occasionally gestures may include head movements or facial expressions. The tasks of the challenge will be "single signer", "small vocabulary" recognition of short continuous sequences of signs. The participants will bring their own hardware and software to the site of the challenge, where they will have to demonstrate their system. We are planning to hold the live evaluation in conjunction with CVPR 2012 (pending acceptance of our proposal). We also intend to run a milestone event at ICCV 2011 at the demo session of the 6th IEEE Workshop on Human Computer Interaction: Real Time Vision Aspects of Natural User Interfaces (pending acceptance of our proposal).