Projects

Video Quality Assessment using Space-Time chips

Developed a new feature space and algorithm for video quality.

Conditional Wasserstein Generative Adversarial Networks for Image Dehazing

Proposed a novel application of CWGANs.

FPGA implementation of fog-removal using anisotropic diffusion.

Implemented and optimized a fog-removal algorithm on an FPGA.

A Survey of Combinatorial Auctions

Performed an extensive survey of Combinatorial Auctions.

Deep learning for skin lesion segmentation

Proposed novel pre-processing techniques to improve identification of cancerous regions in the skin with deep learning.

Online-learning based video reconstruction for adaptive bitrate video streaming

Used the temporal self-similarity in streaming videos to reconstruct frames sent when the channel is poor.

Peer-to-Peer Live Streaming with Recommender Systems

Studied optimal load distributions and asymptotic behaviour of P2P systems when used with recommenders.