Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
define up-to-date all-in-one model for pascal finetuning
load transform params in window data layer
include WindowDataLayer in data param upgrade
Fix typo in LRN-expression in docs
|
|
|
|
|
|
[fix] set R-CNN fine-tuning PASCAL example paths
|
|
|
|
|
|
|
|
Add contrastive loss layer, tests, and a siamese network example
|
|
|
|
|
|
|
|
|
|
|
|
|
|
example using shared weights and the contrastive loss.
|
|
|
|
[example] HDF5 classification
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Not lenet.prototxt
|
|
This makes the readme file more consistent
|
|
|
|
|
|
|
|
|
|
better performance) and terminate at iter 65K
|
|
convention
|
|
test_iter (should be 100 instead of 50)
|
|
(euclidean) error
|
|
MNIST example
|
|
|
|
|
|
|
|
|