- Research Program Mentor
MS Master of Science candidate
Machine learning, deep learning, image processing
Blind deconvolution is the problem of restoring an input signal without any prior knowledge of the system which it passes through. Particularly in images, this means we would like to restore an image without knowing how the blur (or noise in general) is characterized. An interesting project could be applying machine learning to solve this problem, possibly by learning the priors based on training data.