Please browse our papers and other related works below. 

Publications on clustering and classification via lossy coding:

• Classification via Minimum Incremental Coding Length (MICL) [pdf (tech report)] [bibtex] [code]. John Wright, Yangyu Tao, Zhouchen Lin, Yi Ma, Heung-Yeung Shum. accepted by Neural Information Processing Systems (NIPS), 2007. 

• Natural Image Segmentation with Adaptive Texture and Boundary Encoding [pdf] [bibtex] [code] [website]. Shankar R. Rao, Hossein Mobahi, Allen Y. Yang, S. Shankar Sastry, Yi Ma. Proceedings of Asian Conference on Computer Vision (ACCV) 2009, H. Zha, R.-i. Taniguchi, and S. Maybank (Eds.), Part I, LNCS 5994, pp.135--146, Springer. Recepient of Best Student Paper Award: Sang Uk Lee Award (News Coverage at CSL).

• Unsupervised Segmentation of Natural Images via Lossy Data Compression [pdf] [bibtex] [code] [website]. Allen Y. Yang, John Wright, S. Shankar Sastry, Yi Ma. accepted by CVIU, 2007.

• Segmentation of Multivariate Mixed Data via Lossy Coding and Compression [pdf] [bibtex] [code]. Yi Ma, Harm Derksen, Wei Hong and John Wright. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007. 

Related Work and Resources:

• MDL on the Web [website]. Excellent online resource for classical Minimum Description Length (MDL). The approach outlined here differs from most MDL work in that we use lossy coding, but the two techniques have much in common philosophically.