Nilanjan Ray

Professor
Department of Computing Science
University of Alberta, Canada


Image Classification Using Dictionary Learning


Fig. 1: Joint projection and low rank dictionary learning(see publications)

Image classification is a benchmark task in computer vision. We have approached this problem with dictionary learning. Some significant publications came of this (see below). In one of the latest methods, we learning low dictionaries with joint projection and graph constraints (Fig. 1) to solve image classification where labeled data is limited. Our method yielded state-of-the-art accuracy on several benchmark datasets.

Related Publications
  • H. Foroughi, N. Ray, H. Zhang, “Object classification with joint projection and low-rank dictionary learning,” IEEE Transactions on Image Processing, 2017, doi: 10.1109/TIP.2017.2766446
  • H. Foroughi, M. Shakeri, N. Ray, H. Zhang, “Face Recognition Using Multi-modal Low-rank Dictionary Learning,” IEEE ICIP 2017.
  • H. Foroughi, M. Sakeri, N. Ray, H. Zhang, ″Joint feature selection with low-rank dictionary learning,″ BMVC 2015, oral presentation (7% acceptance rate for oral presentation).