114
Voted
NIPS
2008
15 years 4 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...