75
Voted
NIPS
2008
14 years 11 months ago
2008
Prior work has shown that features which appear to be biologically plausible as well as empirically useful can be found by sparse coding with a prior such as a laplacian (L1) that...