Commit Graph

733 Commits

Author SHA1 Message Date
Brandon Amos 0574a72641 Initial resnet model definition. 2016-06-19 17:13:21 -04:00
Brandon Amos 2c6ddfba2d Training: Model: Use opt.imgDim. 2016-06-16 20:12:21 -04:00
Brandon Amos 5132cbf3e7
Update citation to our tech report. 2016-06-14 17:13:03 -04:00
Brandon Amos 1fb8abc247 Training: Improve -data description. 2016-06-14 15:56:37 -04:00
Brandon Amos 92684ab771 Training: Add cuda device option. 2016-06-14 15:55:46 -04:00
Brandon Amos a1b3251f96 Actually fix nans from #127. Error if they appear. 2016-06-14 13:51:18 -04:00
Brandon Amos 2bad4e503e Training Tests: Specify -save. 2016-06-13 14:22:59 -04:00
Brandon Amos c04386c48c Training: Only convert to cuda in cuda mode. 2016-06-13 14:12:36 -04:00
Brandon Amos f4567e882e Fixes for travis. 2016-06-13 14:06:12 -04:00
Brandon Amos fea2fb41d8 Resolve nan issues for #127 (hopefully) 2016-06-13 13:50:38 -04:00
Brandon Amos aaab8fe119 Training: Update plot-loss to new directory structure. 2016-06-13 13:50:38 -04:00
Brandon Amos c6f0962b6d Training: Add -save option. Default to a timestamp instead of numbers. 2016-06-13 13:50:38 -04:00
Brandon Amos 5e0ab14049 align_dlib: Add returns. 2016-06-10 11:05:20 -04:00
Brandon Amos bb2ef270c1 Formatting. 2016-06-09 15:14:26 -04:00
Brandon Amos 252c0c9d43 Fix pydoc for align_dlib.align 2016-06-09 14:53:20 -04:00
Brandon Amos 54a5c9b59f Add align_dlib.align for backwards compatibility. 2016-06-09 14:45:39 -04:00
Brandon Amos e534fa9c7f Initial commit of VGG model definitions. 2016-06-08 12:13:27 -04:00
Brandon Amos 288fd8a3bd Fix test for changed cudnn option. 2016-06-07 18:34:29 -04:00
Brandon Amos bced2bb445 Disable cudnn by default for #127. Add warning. 2016-06-07 18:20:46 -04:00
Brandon Amos 21f1bcc5c4 Temporarily remove optimize_net for #127. 2016-06-07 18:14:13 -04:00
Brandon Amos 40f124c614 Update website to align_v1. 2016-06-07 17:02:16 -04:00
Brandon Amos db46f273d2
Fix API doc default arguments. 2016-06-07 16:53:22 -04:00
Brandon Amos 7800460ad7
Add numpy.linalg as a mock module for API docs.
\cc @Godrek
2016-06-07 16:49:35 -04:00
Brandon Amos fe387e175d Merge pull request #147 from Godrek/align-dlib-v2
Align dlib v2
2016-06-07 16:39:53 -04:00
Dante Knowles 3e679b9fb9 Added align dlib v2 2016-06-07 20:26:28 +00:00
Brandon Amos 49fffb3671 Correct inception tower sizes for #142. 2016-06-03 19:04:40 -04:00
Brandon Amos 130edcecc9 Training: Use ADAM instead of adadelta. 2016-06-03 17:48:38 -04:00
Brandon Amos 02b85b7824
Update SP. 2016-05-25 12:47:47 -04:00
Brandon Amos 815b471b50
Update SP. 2016-05-24 04:22:43 -04:00
Brandon Amos d720f012c3
Fix typo.
Thanks @Thetoxicarcade!
2016-05-20 12:21:47 -04:00
Brandon Amos 3b9b99c3b2 Bump scikit-learn requirements for #137. 2016-05-13 18:24:17 -04:00
Brandon Amos 5a7e265086 Bump Dockerfile dependency for TDS for #132. 2016-05-08 12:15:30 -04:00
Brandon Amos b4df1f9644 Revert dataset.lua for #132. 2016-05-08 12:11:59 -04:00
Brandon Amos 0044a7817c Add new TDS dependency to docs.
\cc @aaronnech

For #132
2016-05-07 20:46:49 -04:00
Brandon Amos 58ac0cf7f0 Add new TDS dependency to the Dockerfile.
\cc @aaronnech

For #132
2016-05-07 20:45:49 -04:00
Brandon Amos 0707db5718 Fix maxPathLength assertion/increment.
Thanks for the correction @aaronnech, I read your earlier suggestion too fast :-)

For #132
2016-05-07 20:43:53 -04:00
Brandon Amos 85d7375fd9 Remove unused variables/functions to fix luacheck warnings.
\cc @aaronnech

For #132
2016-05-07 20:33:52 -04:00
Brandon Amos b11910b1da Fix assert bound.
\cc @aaronnech

For #132
2016-05-07 20:30:17 -04:00
Brandon Amos 1d5491a1e8 Add @aaronnech's dataset.lua improvements for #132. 2016-05-07 20:25:27 -04:00
Brandon Amos 4c0522f651 Docker: Bump dependency version. Remove src.
For #133.
2016-05-06 11:13:19 -04:00
Brandon Amos 9c4fd0f976 Remove OpenCV and dlib source files for #133. 2016-05-06 11:11:58 -04:00
Brandon Amos ecfb83fdb0 Add building/pushing comment to opencv-dlib-torch. 2016-05-06 10:14:16 -04:00
Brandon Amos 60203bd45e Merge pull request #131 from bakwc/master
Added ability to skip images with many faces; fixed threshold type
2016-05-03 19:23:35 -04:00
Filipp Bakanov 30db42c2ba Fixed style-guide 2016-05-03 22:54:01 +03:00
Filipp Bakanov d18017a195 Added ability to skip images with more than one face; fixed threshold type 2016-05-03 22:15:57 +03:00
Brandon Amos 5b46a5fcb4 Fix typo. 2016-04-27 17:30:14 -04:00
Brandon Amos 8423c91364 Mention optnet is optional and only for DNN training.
#126
2016-04-19 14:43:24 -04:00
Brandon Amos b9f41dcb02 Merge pull request #126 from neuraldomi/master
Update setup.md
2016-04-19 14:41:11 -04:00
neuraldomi caaaee5be1 Update setup.md
added 'optnet'
2016-04-19 20:35:34 +02:00
Brandon Amos 88cd4af209
Website: Mention the parameter counts don't include batch normalization means and variances for #125. 2016-04-18 20:25:34 -04:00