Machine Learning Meetings and Events
Group Meetings: Group meetings are held Mondays from 11am to Noon (talk starts 11:10am) in D.L. Pratt 290C unless otherwise noted. Meetings are coordinated by Hugo Larochelle.
Tea Talks: Tea talks are held every Wednesday at 4:00pm in D.L. Pratt 290C. Talks should be simple, accessible, and not exceed 15 minutes. Speakers bring snacks, make tea, and provide a copy of the presented paper.
Tea Talk Aug 12, 2009: Emergence of complex cell properties by learning to generalize in natural scenes, by Yan Karklin & Michael S. Lewicki
- Speaker: Hugo Larochelle
- Link: http://www.nature.com/nature/journal/v457/n7225/full/nature07481.html
- Description:
- Abstract: This paper describes an approach for modeling natural images where, instead of learning a code determining the mean parameter of the image distribution (like in sparse coding), we learn a code parameterizing the covariance matrix of the image distribution. It is shown that several physiological properties of complex cells in V1 and V2 are mimicked by this modeled when trained on patches of natural images. The model also learns a representation that can discriminate between image patches of different textures.