000f33cab4
lfw experiment: Add argument for the pairs.txt location.
2015-10-10 21:28:41 -04:00
d69935eb43
Ignore mp4 videos.
2015-10-10 21:28:41 -04:00
69586648e6
Refactor: facenet->openface.
2015-10-10 21:28:41 -04:00
245a833113
Training: plot-loss: Show last few values.
2015-10-10 21:28:41 -04:00
3e5d1a67a0
Compare demo: Fix options.
2015-10-10 21:28:41 -04:00
51aaf28668
README: Tweak build instructions.
2015-10-04 02:09:05 -04:00
9389e8c123
compare.py: Bump dlib version.
2015-10-04 02:09:05 -04:00
57f9b89942
Dockerfile: Initial commit.
2015-10-04 02:09:05 -04:00
02990255e3
Workaround solution for #5 : Only use the largest face in the video.
2015-10-03 23:46:43 -04:00
dc7b050423
Web demo: Correctly display name.
2015-10-03 23:40:19 -04:00
b0561510b1
Web demo: Don't train SVM if there is only one person.
2015-10-03 23:40:19 -04:00
b974ce13c9
README: Add performance notes.
2015-10-03 23:23:38 -04:00
9abd1ea545
Tweak README.
2015-10-03 23:04:02 -04:00
818d1cd13b
TorchWrap: Use a random hex string in the temporary filename.
2015-10-03 22:59:40 -04:00
dbd4490536
Web demo: Make the 'unknown' class optional.
2015-10-03 22:52:58 -04:00
f12d181a75
Add command-line opts to the web demo.
2015-10-03 22:47:35 -04:00
f51d02fb51
Training: Plot: Set alpha on the test loss since it overlaps the train loss.
2015-10-03 22:47:27 -04:00
4509fac2e3
Fix link and add paragraph to the README.
2015-10-03 22:42:05 -04:00
90ebfd6b8f
Fix YouTube link.
2015-10-03 21:28:50 -04:00
41268e57c5
README: Move sections. Add screenshot and link to YouTube video.
2015-10-03 21:26:59 -04:00
1f7c708f30
Pass on README.
2015-10-03 21:15:36 -04:00
0f077cad53
Add cloc script.
2015-10-03 19:24:22 -04:00
874401e0a6
TorchWrap: Add simple function to write image to disk.
2015-10-03 19:23:59 -04:00
402b453f05
Comparison demo: Use all combinations of images.
2015-10-03 19:23:13 -04:00
28891143a1
Add example images of Lennon and Clapton.
2015-10-03 19:22:55 -04:00
4c91109c40
Add t-SNE script.
2015-10-03 18:46:29 -04:00
aa0c8c5664
Add tsne plots.
2015-10-03 18:43:18 -04:00
aa6ee28120
Training: Loss: log-scale y axis.
2015-10-03 17:24:17 -04:00
1d58e82993
LFW: Show AUC.
2015-10-03 17:16:44 -04:00
718ed8caa1
Add OpenBR v1.1.0 data.
2015-10-03 17:16:02 -04:00
5aaea9cb58
Remove NaN values from eigenfaces data.
2015-10-03 17:15:46 -04:00
53ab6943d4
README: typo.
2015-09-29 18:43:53 -04:00
a4afe1c515
Refactor: examples->demos.
2015-09-29 18:33:43 -04:00
312beca814
Refactor: examples->demos.
2015-09-29 18:32:59 -04:00
4067ba6d2a
Tweak README.
2015-09-29 18:28:37 -04:00
1b09261303
README: Add note. Draft of manual build instructions.
2015-09-29 18:27:36 -04:00
513b498449
Add file comparison example.
2015-09-29 18:07:12 -04:00
875ac72039
Lua Server: Use model in evaluation mode.
2015-09-29 18:03:53 -04:00
429452c806
Python: naive_dlib: Take full paths on input.
2015-09-29 18:03:27 -04:00
06a3afa5d6
Python library: Kill Lua subprocess on exit.
2015-09-29 18:03:00 -04:00
158ff1bcce
LFW: Use 4 decimals. Rename our model.
2015-09-29 18:02:36 -04:00
8728bae648
Batch Represent: Remove unnecessary CUDA dependency.
2015-09-29 18:02:01 -04:00
5bffd697cc
LFW experiment: Show avg and stdev of accuracies.
2015-09-26 04:41:48 -04:00
d85ee28f93
Fully implement the LFW experiment.
...
+ Use ROC curve on 10 folds of test data and average them.
+ Add more ROC curves to the plot for comparison.
+ Compute accuracy from best threshold on the training data.
2015-09-26 04:34:58 -04:00
02c22ecd64
TorchWrap/Server: Add comments linking to issues.
2015-09-25 14:57:17 -04:00
d4830e84fc
TorchWrap: Load model relative to file location. Remove unused parameter.
2015-09-25 14:12:38 -04:00
2f25d3d6e3
Web Demo: Use grid to train SVM. Disable training mode by default.
2015-09-24 18:05:19 -04:00
7d9af6cab1
Pass on README.
2015-09-24 17:36:30 -04:00
2fc6b0e3de
gitignore: Web example files.
2015-09-24 17:36:30 -04:00
81dbe0e595
Initial commit of web demo.
2015-09-24 17:36:30 -04:00
Brandon Amos