Sumit Chakravarty's areas of research:

1.  Application of  Computer Vision for Hyperspectral Image Analysis: In this area we utilize computer vision concepts  for exploitation of Hyperspectral Imagery  in applications like tracking, segmentation and component decomposition

2. Deep Learning for Hyperspectral data exploitation: Deep learning is a new avatar of neural Networks. it has shown remarkable success in applications  like Google Image search and Facebook friend search. This research aims apply deep learning to Hyperspectral remote sensing exploitation. This work is being done in collaboration with the Computer Science Department at Kennesaw.

3. Use of Hyperspectral remote sensing for water quality and other geophysical parameter estimation:  Using GIS and other remote sensing data, the prediction of  water  quality is attempted.

4. Application of  constraint optimization techniques for Compressed Sensing of MRI. Compressed sensing and sparse representations are supper new concepts in the domain of machine learning and computer vision. One of the key applications is in data compression of Magnetic Resonance Imaging data. This avenue of research pursues fast and efficient approaches to enable such applications.

5. Underwater Acoustic Communications: Underwater communications is a challenging domain. it typically uses  acoustics . This research focuses on estimation and prediction of such acoustic channels for efficient communications. It also explores how advanced concepts like cognitive radio and cooperative networks be used effectively in such environment.