No-Reference Image Quality Assessment Using Sparse Representations

Friday, December 16, 2016
UTA 7.532

Abstract: In this talk, I will present a novel blind image quality assessment (BIQA) algorithm inspired by the sparse representation of natural images in the human visual system (HVS). The hypothesis behind the proposed method is that the properties of natural images that afford their sparse representation are altered in the presence of distortion. The change in sparsity is quantified to show that it is indeed a measure of the unnaturalness or distortion in an image. Two methods for this quantification will be discussed - one based on the l 2 norm of the error, and the other based on likelihood estimation. The proposed method delivers competitive performance with the state-of- the-art methods. It is both opinion-unaware and distortion-unaware in addition to generating a distortion map to help localise distortions in an image.


Department of Electrical Engineering, IIT-Hyderabad

Sumohana received his PhD degree in ECE from the University of Texas at Austin in 2007. He is currently Associate Professor of Electrical Engineering at the Indian Institute of Technology Hyderabad. Sumohana received the Excellence in Teaching award at IIT Hyderabad in 2013. His research interests include image and video quality assessment, biomedical imaging and image processing.