|
|
| Results on modeling natural image patches |
|
| using a stack of RBM’s (Osindero and Hinton) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| • | Stack of RBM’s
learned one at a time. |
||||||||||||||||
| • | 400 Gaussian
visible units that see |
||||||||||||||||
| whitened image
patches |
|||||||||||||||||
| – | Derived
from 100,000 Van Hateren |
||||||||||||||||
| image
patches, each 20x20 |
|||||||||||||||||
| • | The hidden units
are all binary. |
||||||||||||||||
| – | The lateral connections are |
||||||||||||||||
| learned
when they are the visible |
|||||||||||||||||
| units
of their RBM. |
|||||||||||||||||
| • | Reconstruction
involves letting the |
||||||||||||||||
| visible units of
each RBM settle using |
|||||||||||||||||
| mean-field
dynamics. |
|||||||||||||||||
| – | The
already decided states in the |
||||||||||||||||
| level
above determine the effective |
|||||||||||||||||
| biases
during mean-field settling. |
|||||||||||||||||
|
|
|
| 1000 top- |
||
| level units. |
||
| No MRF. |
||
