VARSeR (Visual Activity Recognition with SEmantic Reasoning) is a proposition conducted by Vincent Vassout and evaluated by myself. It aims to improve the robot’s vision-based recognition of user gesture by taking into consideration context data from the surrounding smart environment.

Publication

• RAMOLY, N., VASSOUT, V., BOUZEGHOUB, A., EL YACOUBI, M. A., AND HARIZ, M. Refining visual activity recognition with semantic reasoning. In Advanced Information Networking and Applications (AINA), 2017 IEEE 31st International Conference on (2017), IEEE, pp. 720–727