Multiple Dictionary Learning for Blocking Artifacts Reduction


We present a structured dictionary learning method to remove blocking artifacts without blurring edges or making any assumption over image gradients. Instead of a single over complete dictionary, we build multiple subspaces and impose sparsity on nonzero reconstruction coefficients when we project a given texture sample on each subspace separately. In case the texture matches to the data set with which the subspace is trained, the corresponding response will be stronger and that subspace will be chosen to represent the texture. In this manner we compute the representations of all patches in the image and aggregate these to obtain the final image. Since the block artifacts are small in magnitude in comparison to actual image edges, aggregation efficiently removes the artifacts but keep the image gradients. We discuss the choices of subspace parameterizations and adaptation to given data. Our results on a large data set of benchmark images demonstrate that the presented method provides superior results in terms of pixel-wise (PSNR) and perceptual (SSIM) measures.


  • Related News & Events

    •  NEWS    ICASSP 2012: 8 publications by Petros T. Boufounos, Dehong Liu, John R. Hershey, Jonathan Le Roux and Zafer Sahinoglu
      Date: March 25, 2012
      Where: IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
      MERL Contacts: Dehong Liu; Jonathan Le Roux; Petros T. Boufounos
      • The papers "Dictionary Learning Based Pan-Sharpening" by Liu, D. and Boufounos, P.T., "Multiple Dictionary Learning for Blocking Artifacts Reduction" by Wang, Y. and Porikli, F., "A Compressive Phase-Locked Loop" by Schnelle, S.R., Slavinsky, J.P., Boufounos, P.T., Davenport, M.A. and Baraniuk, R.G., "Indirect Model-based Speech Enhancement" by Le Roux, J. and Hershey, J.R., "A Clustering Approach to Optimize Online Dictionary Learning" by Rao, N. and Porikli, F., "Parametric Multichannel Adaptive Signal Detection: Exploiting Persymmetric Structure" by Wang, P., Sahinoglu, Z., Pun, M.-O. and Li, H., "Additive Noise Removal by Sparse Reconstruction on Image Affinity Nets" by Sundaresan, R. and Porikli, F. and "Depth Sensing Using Active Coherent Illumination" by Boufounos, P.T. were presented at the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP).