| a575f23 | Yasser Souri | 04 September 2017, 16:40:32 UTC | update gitignore | 04 September 2017, 16:40:32 UTC |
| f48d33d | Yasser Souri | 26 November 2016, 11:38:16 UTC | update for sklearn upgrade | 26 November 2016, 11:38:16 UTC |
| 7e864e9 | Yasser Souri | 24 September 2016, 09:24:19 UTC | Merge branch 'master' of github.com:yassersouri/Ghiaseddin | 24 September 2016, 09:24:19 UTC |
| 46631cf | Yasser Souri | 24 September 2016, 09:23:38 UTC | Create LICENSE.md | 24 September 2016, 09:23:38 UTC |
| 9936488 | Yasser Souri | 17 September 2016, 09:28:04 UTC | don't use pca initialization in TSNE | 17 September 2016, 09:28:04 UTC |
| 0828cf4 | Yasser Souri | 17 September 2016, 09:27:45 UTC | l2 normalize embeddings before visualization | 17 September 2016, 09:27:45 UTC |
| ede32ef | Yasser Souri | 08 September 2016, 23:29:15 UTC | add pca initialization to tsne + notebook | 08 September 2016, 23:29:15 UTC |
| 7844f9c | Yasser Souri | 08 September 2016, 23:17:25 UTC | remove embeeeing scripts in the notebook folder | 08 September 2016, 23:17:25 UTC |
| e5b6b13 | Yasser Souri | 08 September 2016, 23:11:27 UTC | make pastalog optional | 08 September 2016, 23:11:27 UTC |
| 5b88845 | Yasser Souri | 08 September 2016, 23:07:29 UTC | add genrate_embedding implementation | 08 September 2016, 23:07:29 UTC |
| 1236407 | Yasser Souri | 08 September 2016, 22:56:53 UTC | remove legacy notebooks | 08 September 2016, 22:56:53 UTC |
| 9f8dbdc | Yasser Souri | 08 September 2016, 21:52:10 UTC | add doc string for `_general_image_preprocess` | 08 September 2016, 21:52:10 UTC |
| 1baedec | Yasser Souri | 08 September 2016, 21:41:15 UTC | make weights optional for creating an extractor this eliminates the need to have the initial pretrained weights for an extractor when trying to load a model | 08 September 2016, 21:41:15 UTC |
| d253b6a | Yasser Souri | 07 September 2016, 20:10:43 UTC | pep 8 cleanup | 07 September 2016, 20:17:32 UTC |
| edf4923 | Yasser Souri | 07 September 2016, 20:10:31 UTC | update readme with results | 07 September 2016, 20:10:31 UTC |
| e8c4017 | Yasser Souri | 07 September 2016, 19:01:17 UTC | update readme with images | 07 September 2016, 19:01:17 UTC |
| 84bd622 | Yasser Souri | 07 September 2016, 11:16:07 UTC | update embedding scripts | 07 September 2016, 11:16:07 UTC |
| ab51da9 | Erfan Noury | 06 September 2016, 22:42:15 UTC | Fix import in embeddingprep | 06 September 2016, 22:42:15 UTC |
| 863221e | Erfan Noury | 06 September 2016, 22:41:59 UTC | Fix model root for embeddingprep-lfw | 06 September 2016, 22:41:59 UTC |
| 72550fa | Erfan Noury | 06 September 2016, 22:34:44 UTC | Add embeddingprep for LFW10 | 06 September 2016, 22:34:44 UTC |
| 560e396 | Erfan Noury | 06 September 2016, 22:32:13 UTC | Fix bugs for embeddingprep Zap | 06 September 2016, 22:32:13 UTC |
| 0657389 | Erfan Noury | 06 September 2016, 22:24:01 UTC | Fix typo (adress -> address) | 06 September 2016, 22:24:01 UTC |
| 9ba9056 | Erfan Noury | 06 September 2016, 21:50:44 UTC | Merge branch 'master' of https://github.com/yassersouri/ghiaseddin | 06 September 2016, 21:50:44 UTC |
| 40fd9f6 | Erfan Noury | 06 September 2016, 21:48:13 UTC | Add embeddingprep code for Zappos | 06 September 2016, 21:48:13 UTC |
| 76c365f | Yasser Souri | 06 September 2016, 20:54:12 UTC | fix bug | 06 September 2016, 20:58:07 UTC |
| dced1ff | Yasser Souri | 06 September 2016, 20:04:06 UTC | add script to generate saliencies for the paper | 06 September 2016, 20:04:06 UTC |
| 3536295 | Yasser Souri | 06 September 2016, 19:37:43 UTC | smaller saliency images | 06 September 2016, 19:37:43 UTC |
| 094b785 | Erfan Noury | 06 September 2016, 16:11:39 UTC | Change the saliency display function | 06 September 2016, 16:11:39 UTC |
| cccd34a | Yasser Souri | 06 September 2016, 07:29:54 UTC | update run scripts of osr and pubfig | 06 September 2016, 07:29:54 UTC |
| afdd04a | Yasser Souri | 05 September 2016, 13:16:53 UTC | add zappos2 | 05 September 2016, 13:16:53 UTC |
| 2637bd6 | Yasser Souri | 04 September 2016, 21:31:16 UTC | bug fix | 04 September 2016, 21:31:16 UTC |
| 26dd406 | Yasser Souri | 04 September 2016, 20:03:06 UTC | update ranker and small_train | 04 September 2016, 20:03:06 UTC |
| 9c47b39 | Yasser Souri | 03 September 2016, 10:56:46 UTC | add sampling for OSR test | 03 September 2016, 10:56:46 UTC |
| e65ce14 | Yasser Souri | 02 September 2016, 22:44:16 UTC | add another script for osr and pubfig since the test set is large, I skip some parts of the original train script | 02 September 2016, 22:44:16 UTC |
| c6f76a1 | Yasser Souri | 02 September 2016, 12:17:40 UTC | bug fix | 02 September 2016, 12:17:40 UTC |
| 8d662ae | Yasser Souri | 02 September 2016, 12:07:57 UTC | add scripts to run osr and pubfig | 02 September 2016, 12:07:57 UTC |
| c1a2f42 | Yasser Souri | 02 September 2016, 06:07:34 UTC | print the random seed in the train script | 02 September 2016, 06:07:34 UTC |
| 4ce42e1 | Yasser Souri | 01 September 2016, 12:10:37 UTC | once again, bug fix | 01 September 2016, 12:10:37 UTC |
| a595acd | Yasser Souri | 01 September 2016, 12:08:23 UTC | bug fix | 01 September 2016, 12:08:23 UTC |
| dd034ca | Yasser Souri | 01 September 2016, 11:04:19 UTC | make baseline and train scripts one also change the scripts to use the new one also add extractor parameter to train script | 01 September 2016, 11:04:19 UTC |
| 799544a | Yasser Souri | 01 September 2016, 10:47:41 UTC | train with vgg | 01 September 2016, 10:47:41 UTC |
| 2fbdfcd | Yasser Souri | 31 August 2016, 21:37:47 UTC | better log location | 31 August 2016, 21:37:47 UTC |
| 1095f74 | Yasser Souri | 31 August 2016, 21:37:36 UTC | fix zappos bug | 31 August 2016, 21:37:36 UTC |
| 0a28542 | Yasser Souri | 31 August 2016, 15:45:24 UTC | remove debug line | 31 August 2016, 15:45:24 UTC |
| 3327ffe | Yasser Souri | 31 August 2016, 15:44:21 UTC | Merge pull request #9 from yassersouri/augmentation Dataset Augmentation | 31 August 2016, 15:44:21 UTC |
| 98c4027 | Yasser Souri | 31 August 2016, 15:42:34 UTC | update notebook | 31 August 2016, 15:42:34 UTC |
| e9b1ce6 | Yasser Souri | 31 August 2016, 15:41:41 UTC | add augmentation also add optional logging and fix a bug in train_n_iter | 31 August 2016, 15:41:41 UTC |
| d8cac8e | Yasser Souri | 31 August 2016, 12:00:45 UTC | Merge branch 'augmentation' of github.com:yassersouri/Ghiaseddin into augmentation Conflicts: ghiaseddin/datasets.py notebooks/saliency_playground.ipynb | 31 August 2016, 12:00:45 UTC |
| 7f38eb6 | Yasser Souri | 31 August 2016, 11:50:07 UTC | refactor random augmentation methods to a better place It is more natural to have the extractor know and handel augmentaion. | 31 August 2016, 11:50:07 UTC |
| 2deda54 | Yasser Souri | 31 August 2016, 11:02:38 UTC | remove unnecessary import | 31 August 2016, 11:47:54 UTC |
| 6a348f0 | Yasser Souri | 31 August 2016, 10:56:02 UTC | add basic augmentation in dataset class add a way to visualize the augmentation | 31 August 2016, 11:47:54 UTC |
| 73876e8 | Yasser Souri | 31 August 2016, 10:29:39 UTC | naming of dataset when augmentation is on | 31 August 2016, 11:47:54 UTC |
| e1f66e9 | Yasser Souri | 31 August 2016, 11:44:45 UTC | update readme with scripts to download initial weights | 31 August 2016, 11:44:45 UTC |
| 5b1b82c | Yasser Souri | 31 August 2016, 10:56:02 UTC | add basic augmentation in dataset class add a way to visualize the augmentation | 31 August 2016, 10:56:02 UTC |
| 3c497e5 | Yasser Souri | 31 August 2016, 10:29:39 UTC | naming of dataset when augmentation is on | 31 August 2016, 10:29:39 UTC |
| 7faa269 | Yasser Souri | 31 August 2016, 09:15:47 UTC | update readme | 31 August 2016, 09:15:47 UTC |
| dd3ffeb | Yasser Souri | 31 August 2016, 09:08:02 UTC | Merge branch 'debug-trained-model' | 31 August 2016, 09:08:02 UTC |
| 0c49405 | Yasser Souri | 31 August 2016, 09:06:18 UTC | Merge branch 'debug-trained-model' of github.com:yassersouri/Ghiaseddin into debug-trained-model | 31 August 2016, 09:06:18 UTC |
| f71d9d3 | Yasser Souri | 31 August 2016, 08:48:52 UTC | update notebook | 31 August 2016, 09:04:55 UTC |
| c42d2e1 | Yasser Souri | 30 August 2016, 23:13:14 UTC | fix bug | 31 August 2016, 09:04:55 UTC |
| bf08777 | Yasser Souri | 30 August 2016, 23:03:55 UTC | add saliency maps to the train script | 31 August 2016, 09:04:54 UTC |
| aaae3b1 | Yasser Souri | 30 August 2016, 22:44:21 UTC | fix bug | 31 August 2016, 09:04:54 UTC |
| 10e2848 | Yasser Souri | 30 August 2016, 22:33:21 UTC | update notebook with saliency example | 31 August 2016, 09:04:54 UTC |
| 6123c94 | Yasser Souri | 30 August 2016, 22:32:34 UTC | add code for saliency generation | 31 August 2016, 09:04:54 UTC |
| 480b7a5 | Yasser Souri | 30 August 2016, 22:31:13 UTC | update scripts | 31 August 2016, 09:04:54 UTC |
| 685dfde | Yasser Souri | 30 August 2016, 22:30:38 UTC | load error handeling | 31 August 2016, 09:04:54 UTC |
| 9a6294f | Yasser Souri | 30 August 2016, 20:27:01 UTC | update notebooks | 31 August 2016, 09:04:54 UTC |
| 5e72658 | Yasser Souri | 30 August 2016, 20:26:10 UTC | improve train script also add shell script to do training and evaluation for zappos | 31 August 2016, 09:04:54 UTC |
| 9728889 | Yasser Souri | 30 August 2016, 20:25:00 UTC | show gray in graphs when equal on test | 31 August 2016, 09:04:54 UTC |
| 7e3da76 | Yasser Souri | 28 August 2016, 11:17:22 UTC | visualizing the output during training you can see correctness, prediction and raw estimates visualized during training | 31 August 2016, 09:04:54 UTC |
| b88aa29 | Yasser Souri | 27 August 2016, 21:40:44 UTC | if nan loss, then do warning also prevent pastalog from throwing exceptions | 31 August 2016, 09:04:54 UTC |
| 1dc7ff3 | Yasser Souri | 27 August 2016, 21:13:12 UTC | formatting change | 31 August 2016, 09:04:54 UTC |
| 9a447be | Yasser Souri | 27 August 2016, 21:12:50 UTC | better conv1 saving | 31 August 2016, 09:04:54 UTC |
| f4cdcf6 | Yasser Souri | 27 August 2016, 21:00:59 UTC | generate conv1 filters show in notebook and save to file | 31 August 2016, 09:04:54 UTC |
| db7dcde | Yasser Souri | 27 August 2016, 20:27:31 UTC | fix some bugs resulted by new changes | 31 August 2016, 09:04:54 UTC |
| f799a45 | Yasser Souri | 27 August 2016, 19:33:21 UTC | add comments | 31 August 2016, 09:04:54 UTC |
| c005222 | Yasser Souri | 26 May 2016, 15:51:51 UTC | add generating error os the model and other improvements | 31 August 2016, 09:03:23 UTC |
| ff13d51 | Yasser Souri | 26 May 2016, 11:35:55 UTC | add skeleton code | 31 August 2016, 09:03:23 UTC |
| e764937 | Yasser Souri | 31 August 2016, 08:48:52 UTC | update notebook | 31 August 2016, 08:48:52 UTC |
| 3ca2c5f | Yasser Souri | 30 August 2016, 23:13:14 UTC | fix bug | 30 August 2016, 23:13:14 UTC |
| d079014 | Yasser Souri | 30 August 2016, 23:03:55 UTC | add saliency maps to the train script | 30 August 2016, 23:03:55 UTC |
| 28f6137 | Yasser Souri | 30 August 2016, 22:44:21 UTC | fix bug | 30 August 2016, 22:44:21 UTC |
| e431487 | Yasser Souri | 30 August 2016, 22:33:21 UTC | update notebook with saliency example | 30 August 2016, 22:33:21 UTC |
| f2e8201 | Yasser Souri | 30 August 2016, 22:32:34 UTC | add code for saliency generation | 30 August 2016, 22:32:34 UTC |
| f2c9b8f | Yasser Souri | 30 August 2016, 22:31:13 UTC | update scripts | 30 August 2016, 22:31:13 UTC |
| 4357ea3 | Yasser Souri | 30 August 2016, 22:30:38 UTC | load error handeling | 30 August 2016, 22:30:38 UTC |
| 66ebbfa | Yasser Souri | 30 August 2016, 20:27:01 UTC | update notebooks | 30 August 2016, 20:27:01 UTC |
| 50e4436 | Yasser Souri | 30 August 2016, 20:26:10 UTC | improve train script also add shell script to do training and evaluation for zappos | 30 August 2016, 20:26:10 UTC |
| a391e47 | Yasser Souri | 30 August 2016, 20:25:00 UTC | show gray in graphs when equal on test | 30 August 2016, 20:25:00 UTC |
| 8408a1b | Yasser Souri | 28 August 2016, 11:17:22 UTC | visualizing the output during training you can see correctness, prediction and raw estimates visualized during training | 28 August 2016, 11:17:22 UTC |
| 2488e49 | Yasser Souri | 27 August 2016, 21:40:44 UTC | if nan loss, then do warning also prevent pastalog from throwing exceptions | 27 August 2016, 21:40:44 UTC |
| 57110b0 | Yasser Souri | 27 August 2016, 21:13:12 UTC | formatting change | 27 August 2016, 21:13:12 UTC |
| 0bd0d7e | Yasser Souri | 27 August 2016, 21:12:50 UTC | better conv1 saving | 27 August 2016, 21:12:50 UTC |
| c9326bc | Yasser Souri | 27 August 2016, 21:00:59 UTC | generate conv1 filters show in notebook and save to file | 27 August 2016, 21:00:59 UTC |
| 75f458e | Yasser Souri | 27 August 2016, 20:27:31 UTC | fix some bugs resulted by new changes | 27 August 2016, 20:27:31 UTC |
| 178f71e | Yasser Souri | 27 August 2016, 19:33:21 UTC | add comments | 27 August 2016, 19:33:21 UTC |
| fc86398 | Yasser Souri | 23 August 2016, 08:10:12 UTC | Merge branch 'debug-trained-model' of github.com:yassersouri/Ghiaseddin into debug-trained-model | 23 August 2016, 08:10:12 UTC |
| b7a06a5 | Yasser Souri | 26 May 2016, 15:51:51 UTC | add generating error os the model and other improvements | 23 August 2016, 08:09:46 UTC |
| f07f51c | Yasser Souri | 26 May 2016, 11:35:55 UTC | add skeleton code | 23 August 2016, 08:09:46 UTC |
| 3aab5ba | Yasser Souri | 23 August 2016, 08:05:42 UTC | new features: - shuffle for test generator - also train_n_iter for the ranker | 23 August 2016, 08:05:42 UTC |
