display picture

Rahul Sharma


Final year student of BTech - MTech dual degree
Department of Electrical Engineering
IIT Kanpur - 208016
Phone : +91-8874356323
E-mail : rahus@iitk.ac.in, ra.rahulsharma.sh@gmail.com

About Me

I am a final year Electrical Engineering student pursuing a five year B.Tech-M.Tech dual degree program at Indian Institute of Technology, Kanpur. I work under guidance of Dr. Tanya Guha and Dr. Gaurav Sharma. I am an active member of the groups SiMPL and Computer Vision. I am primarily interested in applying machine learning algorithms to AudioVisual signals in order to understand human actions, emotions and behaviour (renowned as 'Affective and Behavioral Computing'). Further, I would like to apply my understandings in directions of creating better Human-Computer interfaces.

Current Work (Master's Thesis)

Towards Multimodal Assessment of Speaker Performance in Public Speaking : We propose a computational framework for quantifying speaker performance in the context of public speaking. For this purpose, we created a database consisting of more than two thousand videos along with associated metadata. We do not consider the content of the talk, rather analyze the speech and the visual content to capture the verbal and non-verbal behavior of the speaker. Inspired by the celebrated acceptance of deep learning frameworks over conventional methods, we are trying to model the speaker's behaviour using various deep neural architectures.

Publications

  • R. Sharma, T. Guha "A Trajectory Clustering Approach To Crowd Flow Segmentation in Videos", Proceedings of IEEE International Conference on Image Processing, September 2016. [pdf]
  • R. Sharma, T. Guha and Gaurav Sharma "Towards Multimodal Assessment of Speaker Performance in Public Speaking", To Submit.

References

  • Prof. Gaurav Sharma
  • Computer Science and Engineering, IIT Kanpur
    E-mail : grv[at]cse.iitk.ac.in
    URL : http://www.grvsharma.com/