Research

My research interests span the areas of computer vision, robotic perception and machine learning. I am particularly interested in applying statistical models and developing novel machine learning approaches (e.g. deep learning models) for visual perception tasks, such as human behaviour analysis, dynamic scene understanding, etc.

Recent projects

   cnn-multiscale                     

 Learning Deep Representations for Robotic Perception


BMVC2015dan.png         

Discovering Activity Patterns in Video Scenes with Deep Networks


SALSA

Analysis of Social Scenes from Multimodal Data


HR                     

ANALYSIS OF EMOTIONAL & PHYSIOLOGICAL SIGNALS FROM FACES