mirrors
/
deep-person-reid
mirror of https://github.com/KaiyangZhou/deep-person-reid.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
804 Commits
3 Branches
20 Tags
111 MiB
Commit Graph

260 Commits (d60ca7f2241e4aa5ea2cfffce9ee2fde21f0d901)

Author SHA1 Message Date
KaiyangZhou 024913f73c update readme 2019-05-06 10:48:03 +01:00
KaiyangZhou 6fd0ab364f raise error if batch_size is smaller than num_instances 2019-05-06 10:47:48 +01:00
KaiyangZhou d0af8eb891 0.7.4 2019-04-27 23:31:05 +01:00
KaiyangZhou bb6dc46e21 bug fix in combineall 2019-04-27 23:30:32 +01:00
KaiyangZhou a52d1289f8 add attribute _junk_pids to dataset class 2019-04-27 23:02:54 +01:00
KaiyangZhou dcd8da565a bug fix in lr_scheduler 2019-04-26 15:45:23 +01:00
KaiyangZhou 28b4580a0d remove print in internal wrapper 2019-04-23 11:35:22 +01:00
KaiyangZhou dee01229da remove division by 1e+06 when counting model size 2019-04-23 11:34:52 +01:00
KaiyangZhou 4b93121496 minor 2019-04-19 11:52:02 +01:00
KaiyangZhou 1decc2473a 0.7.2->0.7.3 2019-04-18 12:10:54 +01:00
KaiyangZhou 1ea7787b6e typo fix 2019-04-17 09:13:04 +01:00
KaiyangZhou 4a033659b0 do not decay lr while fixbase=True 2019-03-29 11:40:23 +00:00
KaiyangZhou 4cf18f0830 print domain info while doing evaluation 2019-03-27 22:30:08 +00:00
KaiyangZhou 1e9d466f42 move scheduler.step() to train() 2019-03-27 13:01:39 +00:00
KaiyangZhou 8c1c0084b8 0.7.1 -> 0.7.2 2019-03-25 20:08:31 +00:00
kaiyangzhou eaba4f5cfa rm prid450s 2019-03-25 20:03:43 +00:00
kaiyangzhou 5cc8043c4a verbose=False after __add__ 2019-03-25 20:03:21 +00:00
KaiyangZhou 866744914b 0.7.0 -> 0.7.1 2019-03-25 18:49:20 +00:00
KaiyangZhou 235348f672 bug fix in __add__ 2019-03-25 18:43:19 +00:00
KaiyangZhou 1e7bbc3668 add \n 2019-03-25 16:27:39 +00:00
KaiyangZhou 8e5265099d change weight path 2019-03-25 09:29:06 +00:00
KaiyangZhou e5b8c464ca update docstrings 2019-03-24 23:33:54 +00:00
KaiyangZhou 4ca182dbb8 add model public keys 2019-03-24 15:52:39 +00:00
KaiyangZhou a0e54ede32 add scripts 2019-03-24 11:06:12 +00:00
KaiyangZhou d0fe37c558 finalize docs 2019-03-23 23:09:39 +00:00
KaiyangZhou aa98ef5a7b add docstrings to losses, metrics, models, utils, optim 2019-03-22 00:14:41 +00:00
KaiyangZhou cbbde86eef add docs to torchreid.data 2019-03-21 17:28:14 +00:00
KaiyangZhou 75dd835e42 change batch_size in video 2019-03-21 14:12:50 +00:00
KaiyangZhou 77c47b4e53 add data module 2019-03-21 13:00:07 +00:00
KaiyangZhou 46a0d5c27d update dataset and docs 2019-03-21 12:53:21 +00:00
KaiyangZhou 37f92b7cc3 create engine SDK 2019-03-19 17:26:08 +00:00
KaiyangZhou 03d0674dcb comply woth PEP8 2019-03-15 23:17:38 +00:00
KaiyangZhou 983c9e51e5 add --combineall 2019-03-15 17:45:47 +00:00
KaiyangZhou 402e054110 rearrange order of functions 2019-03-15 16:20:47 +00:00
KaiyangZhou 3cbeab42c2 add init_attributes to BaseDataset 2019-03-15 16:11:25 +00:00
KaiyangZhou e2b1001165 polish code 2019-03-15 14:49:18 +00:00
KaiyangZhou 850da2bbed update path to pretrained weights of resnext 2019-03-12 16:46:49 +00:00
KaiyangZhou 6f3b057c1b add space between class and __init__ (trivial) 2019-03-09 12:53:23 +00:00
KaiyangZhou 5cd0776eb5 rm center loss & ring loss 2019-03-09 12:45:33 +00:00
KaiyangZhou 25c64fdd9a handle UnicodeDecodeError when using python3 2019-03-08 22:27:11 +00:00
KaiyangZhou dc99054af5 compute mAP for cuhk03 old split with multi-gallery-shot 2019-03-07 09:42:06 +00:00
KaiyangZhou 7ff331f871 update mobilenetv2; add imagenet/reid weights 2019-03-05 10:56:39 +00:00
KaiyangZhou 3fe5cb118c polish function name 2019-02-27 09:57:48 +00:00
KaiyangZhou 83b9b38280 update ilids and prid 2019-02-27 09:31:12 +00:00
KaiyangZhou 4078f18aff update print msg 2019-02-20 22:28:30 +00:00
KaiyangZhou 5f971cc03a correct print 2019-02-20 22:22:46 +00:00
KaiyangZhou c427e88658 rank={:.1%} 2019-02-20 22:20:36 +00:00
KaiyangZhou d8cfa82594 add print to save_ckpt 2019-02-20 22:17:47 +00:00
KaiyangZhou 19e03d4097 add remove_module_from_keys option to save_ckpt 2019-02-20 22:13:26 +00:00
KaiyangZhou d56e216068 bug fix 2019-02-20 21:57:09 +00:00
KaiyangZhou 74a2377ab7 add resume_from_ckpt 2019-02-20 21:51:49 +00:00
KaiyangZhou ae54ca0f6b mv save_ckpt to torchtools 2019-02-20 21:51:24 +00:00
KaiyangZhou 26a9400609 change double quotes to single quote in print 2019-02-20 21:50:47 +00:00
kaiyangzhou 356a677cd3 mv basic args to base-manager 2019-02-19 15:48:10 +00:00
kaiyangzhou 1cdb312007 add --color-jitter & --color-aug 2019-02-19 15:44:56 +00:00
kaiyangzhou 859b03cf19 set pretrained=True instead of pretrained="imagenet" 2019-02-19 11:10:45 +00:00
KaiyangZhou 2a5a951666 wrap lr_scheduler 2019-02-08 11:44:12 +00:00
KaiyangZhou 3d67488edd support staged_lr 2019-02-03 14:04:10 +00:00
KaiyangZhou cf5b9a3ba5 update comments 2019-02-03 14:03:48 +00:00
KaiyangZhou bac2ce5662 add random erasing 2019-02-03 14:03:06 +00:00
KaiyangZhou 9c159569fd update author info 2019-02-03 14:01:26 +00:00
KaiyangZhou 043c820d01 ignore "module." when saving/loading weights 2019-02-02 14:20:56 +00:00
KaiyangZhou 760d68ebd1 bug fix 2019-02-01 22:43:47 +00:00
KaiyangZhou 6891a6424a mv model.cuda() before init_optim() 2019-02-01 22:30:41 +00:00
KaiyangZhou 2cb1031942 add load_pretrained_weights() to torchtools 2019-02-01 13:10:23 +00:00
KaiyangZhou 800df43fee check if output is list 2019-02-01 00:50:28 +00:00
KaiyangZhou 564f5a26ab update print 2019-01-30 22:41:47 +00:00
KaiyangZhou b2c06364a3 unify sampler 2019-01-28 18:50:09 +00:00
KaiyangZhou daf184608c add self. 2019-01-28 17:29:22 +00:00
KaiyangZhou f045778602 add accuracy 2019-01-27 23:15:38 +00:00
KaiyangZhou 25aecab2da bug fix 2019-01-24 22:12:02 +00:00
KaiyangZhou 825fce92eb support msmt_v2 2019-01-22 18:18:59 +00:00
kaiyangzhou ef2bacbc00 add market1501-500k 2019-01-22 16:07:43 +00:00
KaiyangZhou 4b6057b923 wrap set_random_seed 2019-01-09 21:30:46 +00:00
KaiyangZhou 10b074e9b7 add os.path.expanduser(root) to dataset files 2019-01-09 13:43:39 +00:00
KaiyangZhou 074a10e24e bug fix 2018-12-27 10:25:09 +00:00
KaiyangZhou f8a31da4c6 replace self.pin_memory with self.use_gpu 2018-12-26 11:10:34 +00:00
KaiyangZhou c05f96e31b add link to shufflenet's imagenet weights 2018-11-26 22:07:20 +00:00
KaiyangZhou 47b883d0d6 bug fix: rm resnext101_32x4d 2018-11-26 22:06:33 +00:00
KaiyangZhou 534dfa6cce update code: add self to image_training 2018-11-24 16:48:07 +00:00
KaiyangZhou 9c373ff6e6 bug fix: add one output for unpacking trainloader 2018-11-24 16:25:10 +00:00
KaiyangZhou affeaa847d add resnext50_32x4d's imagenet weights 2018-11-23 10:31:38 +00:00
KaiyangZhou d7755cfd16 add resnext50_32x4d_fc512 2018-11-22 12:41:22 +00:00
KaiyangZhou b065f130d0 rename tracklet_info to tracklet_stats in datasets/base.py 2018-11-21 09:59:43 +00:00
KaiyangZhou e9a2b7c3d3 add overhead sign 2018-11-12 11:45:41 +00:00
KaiyangZhou d8498ea3dd optimize cython code 2018-11-12 11:44:20 +00:00
KaiyangZhou 8ca8be28c2 update test_cython.py: add cuhk03's comparison 2018-11-11 23:02:20 +00:00
KaiyangZhou 8631603539 rm *.c files for cython 2018-11-11 17:43:28 +00:00
KaiyangZhou 7fe189f00f add torchreid/eval_cylib/test_cython.py 2018-11-10 22:01:25 +00:00
KaiyangZhou 75279f0728 add evaluate_py 2018-11-10 21:25:40 +00:00
KaiyangZhou 41f1f6c6ac fix bug: add return 2018-11-10 21:16:31 +00:00
KaiyangZhou d885f11ce7 update cython code: add cuhk03-metric 2018-11-10 21:09:13 +00:00
KaiyangZhou 1d0851e8c4 fix bug: add self.num_instances to ImageDataManger 2018-11-10 11:59:01 +00:00
KaiyangZhou 8c53007138 return (imgs, pids, camids, img_paths) for image dataloaders 2018-11-10 11:54:06 +00:00
KaiyangZhou db157c1345 update help comments for data_source 2018-11-10 11:53:25 +00:00
KaiyangZhou 0545358b6f mv resnet50mid to reid-specific model list 2018-11-09 16:17:29 +00:00
KaiyangZhou 54b92ee617 check if isinstance(model, nn.DataParallel) 2018-11-08 23:54:26 +00:00
KaiyangZhou 205df55987 add fixbase functions (debug mode) 2018-11-08 23:04:50 +00:00
KaiyangZhou cac6d72a6d minor change: data_source (Dataset) -> data_source (list) 2018-11-08 22:34:24 +00:00
KaiyangZhou 9980b5378b add support for htri training 2018-11-08 22:02:53 +00:00
KaiyangZhou 01fe76cf27 sample -> sample_method 2018-11-08 21:40:44 +00:00
KaiyangZhou e17bbcddf3 source/target -> source_names/target_names 2018-11-08 21:40:18 +00:00
KaiyangZhou c32cf80a9f add RankLogger 2018-11-08 21:38:22 +00:00
KaiyangZhou fdc451c0e5 add linear decay 2018-11-07 22:02:23 +00:00
KaiyangZhou 2d32f5b28a add warnings 2018-11-07 21:55:04 +00:00
KaiyangZhou e3ae601ffc mv KeyError to ValueError 2018-11-07 21:54:46 +00:00
KaiyangZhou 7b47e645e9 add property 2018-11-07 21:46:23 +00:00
KaiyangZhou 1f1f01669b change dataset to (query, gallery) 2018-11-07 21:45:48 +00:00
KaiyangZhou 3f9e20a55e change access to dataloaders in dm by class method 2018-11-07 20:48:21 +00:00
KaiyangZhou e6464f3443 update args for dataset/optimizer 2018-11-07 17:09:23 +00:00
KaiyangZhou b81c13fd10 fix bug: query -> gallery 2018-11-07 15:54:10 +00:00
KaiyangZhou 2ecd8e5704 wrap argument_parser in args.py 2018-11-07 15:36:49 +00:00
KaiyangZhou a96dd6b916 support multi-dataset training 2018-11-05 21:36:12 +00:00
KaiyangZhou 81d3cc3995 fixed division error: / -> // 2018-11-05 21:30:40 +00:00
KaiyangZhou 9726b8bb96 set required=True for dataset arguments 2018-11-05 21:25:09 +00:00
KaiyangZhou 02bac86c71 add data_manager.py 2018-11-05 21:19:27 +00:00
KaiyangZhou e96ac0e3a4 add warning to non-cython evaluation 2018-11-05 19:24:20 +00:00
KaiyangZhou df0ae6d0e0 add num_cams; create base classes 2018-11-04 22:59:46 +00:00
KaiyangZhou 01d7594011 make camid start from 0 2018-11-04 21:27:08 +00:00
KaiyangZhou 39df4855c1 add mlfn 2018-11-04 00:18:02 +00:00
KaiyangZhou 733d967345 rm multi-crop 2018-11-02 11:39:19 +00:00
KaiyangZhou c388c24880 wrap transforms into build_transforms() 2018-11-01 22:29:19 +00:00
KaiyangZhou e302af017d recover dropout in pcb 2018-11-01 17:28:15 +00:00
KaiyangZhou a57558738e generalize CrossEntropyLableSmooth to CrossEntropyLoss 2018-10-29 23:00:17 +00:00
KaiyangZhou 752a4fe498 update pcb: commented dropout 2018-10-28 22:30:19 +00:00
KaiyangZhou 7588c8f581 update pcb: add support to various resnet archs 2018-10-28 11:44:09 +00:00
KaiyangZhou 12cdbf96b5 update pcb: mv conv5 after eval-mode in forward 2018-10-28 11:30:35 +00:00
KaiyangZhou bfba6389d6 add pcb 2018-10-28 11:16:42 +00:00
KaiyangZhou f77b9ada10 fill space after comma 2018-10-27 23:22:23 +01:00
KaiyangZhou bc942aaf09 add variable model_url 2018-10-27 22:50:22 +01:00
KaiyangZhou 3ddf9ce699 build new classifier; add squeezenet1_0_fc512 2018-10-27 22:49:53 +01:00
KaiyangZhou ee5bcee63f add delimiter for lightweight models 2018-10-27 17:59:10 +01:00
KaiyangZhou afca4b442e add delimiter to separate image-classification models and reid models 2018-10-27 17:49:11 +01:00
KaiyangZhou 4ce9024fcc add reference to all models 2018-10-27 17:46:48 +01:00
KaiyangZhou d187902ed4 add resnext 2018-10-27 16:52:05 +01:00
KaiyangZhou 2c2a9c0098 update model code 2018-10-27 14:42:28 +01:00
KaiyangZhou 62aad1324a add import 2018-10-27 14:12:51 +01:00
KaiyangZhou 0af4b88fba add reference 2018-10-27 14:09:21 +01:00
KaiyangZhou 345093a0ba change pretrained=True 2018-10-27 12:01:51 +01:00
KaiyangZhou 35235fe4bb update densenet 2018-10-27 11:57:48 +01:00
KaiyangZhou d076e41e78 add def featuremaps(self, x) 2018-10-27 10:49:27 +01:00
KaiyangZhou fffa92d55c add self.loss 2018-10-27 10:40:20 +01:00
KaiyangZhou b5c305fcce add **kwargs to class __init__() 2018-10-27 10:35:47 +01:00
KaiyangZhou ee5fb676bf add dropout_p; update senet 2018-10-26 23:54:13 +01:00
KaiyangZhou 710b9ac1a2 update model code 2018-10-26 23:05:40 +01:00
KaiyangZhou 1feb4644b9 update split generation in cuhk01 2018-10-10 16:10:26 +01:00
KaiyangZhou 70365320f5 correct split generation in viper 2018-10-08 16:44:26 +01:00
KaiyangZhou 3bd674f198 use imgname as dirname for each query 2018-10-06 22:21:11 +01:00
KaiyangZhou e6ce128a26 update print 2018-10-06 22:16:28 +01:00
KaiyangZhou 86eb3336c9 update print 2018-10-06 22:16:07 +01:00