Sushobhan Ghosh & Sudarshan Nagesh Earn 2nd prize at CVPR 2018 UG2 Challenge

The contest was to design an algorithm to remove artifacts in images captured in challenging conditions such as drones and gliders.

Sushobhan Ghosh & Sudarshan Nagesh

EECS PhD Student Sushobhan Ghosh and Sudarshan Nagesh (Research Specialist in the The Computational Photography Lab) won 2nd prize (12,500$) for the CVPR 2018 UG2 Prize Challenge, held June 18 in Salt Lake City, Utah, organized by IARPA. The contest was to design an algorithm to remove artifacts in images captured in challenging conditions such as drones and gliders.

CVPR is the premier annual computer vision event comprising the main conference and several co-located workshops and short courses. With its high quality and low cost, it provides an exceptional value for students, academics and industry researchers.

The workshop challenge aimed to advance the analysis of images collected by small UAVs by improving the image restoration and enhancement algorithm performance. Ghosh and Nagesh's submission, saw the team design a neural network, which was declared runner-up in the visual enhancement challenge, whose goal was to design an algorithm for enhancing images for classification performance.

Ghosh completed his undergrad at Indian Institute of Technology Delhi (IIT Delhi), India in 2016. Sushobhan is interested in machine learning applications in computational photography and computer vision problems. His current research topics include deep learning based image enhancement in challenging environment, ptychography using automatic differentiation, deep learning based fourier ptychography and time of flight imaging.

Nagesh received the B.Tech. degree from National Institute of Technology Karnataka (NITK), Surathkal, India, in 2012, and the M.Sc. (Engineering) degree in signal processing from Indian Institute of Science (IISc), Bangalore, India in 2015. He received the MS degree in computational imaging from the Rice University, Houston, USA in 2017. His research interests include computational imaging, ptychography, phase-retrieval, and finite-rate-of innovation signal sampling.

Ghosh and Nagesh are members of the Computational Photography Lab, led by Prof. Oliver Cossairt. The lab builds a new breed of cameras with increased functionality and performance that will transform the way cameras are built in coming decades, from low-end consumer imagers to high-end scientific instruments, with research focusing in the emerging field of computational photography, which combines expertise in optics, image processing, computer vision, and computer graphics.