Commit Graph

  • fc6b3d0ad2 fix makefile comment bug oneWayOut 2020-07-06 12:49:57 +0800
  • 76a316cec5 Disable jitter and flip augmentation for contrastive detector learning AlexeyAB 2020-07-06 03:32:14 +0300
  • f13532ce60 Contrastive learning for training Detector to track objects AlexeyAB 2020-07-06 01:50:44 +0300
  • f7c13cf84a Updates README Muhammad Maaz 2020-07-04 17:20:48 +0500
  • e433c4113d
    Merge branch 'master' into softmax Laurynas 2020-06-30 23:33:17 +0100
  • 3af6781de2
    Update Readme.md Alexey 2020-06-30 23:32:13 +0300
  • 320e6fd8d2
    Minor fix Alexey 2020-06-27 16:06:09 +0300
  • 9b3fdb0a07
    Update Readme.md Alexey 2020-06-26 20:19:19 +0300
  • a71b9a6e9a max_delta= affects counters_per_class= AlexeyAB 2020-06-26 01:00:11 +0300
  • dfbf42cfc0
    How to train tiny-yolo Alexey 2020-06-25 17:32:14 +0300
  • dfb3773e7b
    Minor fix Alexey 2020-06-25 17:11:41 +0300
  • c044b98c64
    Minor fix Alexey 2020-06-25 17:04:28 +0300
  • feaa5f62f4
    Update Readme.md Alexey 2020-06-25 17:03:31 +0300
  • 457149dcd1
    Update Readme.md Alexey 2020-06-25 17:02:19 +0300
  • 90004c119d Merge branch 'master' of github.com:AlexeyAB/darknet AlexeyAB 2020-06-25 16:36:00 +0300
  • 918fabccec Yolov4-tiny AlexeyAB 2020-06-25 16:35:36 +0300
  • 5963c22adf Fix '"/*" within comment' warnings; remove instances of '//*' so as to not confuse some syntax highlighters Juuso Alasuutari 2020-06-24 16:20:20 +0300
  • 51d59cf30a
    Merge pull request #6051 from imaami/alexeyab-fix-train_classifier-double-free Alexey 2020-06-24 02:06:50 +0300
  • ad16414923 Fix a double free error in train_classifier() introduced by a merge Juuso Alasuutari 2020-06-24 01:50:45 +0300
  • 17b806f0bd Fixed Loss scale AlexeyAB 2020-06-22 14:50:51 +0300
  • 8f900493c6 SAT for contrastive loss AlexeyAB 2020-06-21 03:02:52 +0300
  • 55d600a002
    fixes license url Iordanis Kostelidis 2020-06-17 23:23:12 +0300
  • 088b7963d3
    fixes the license url Iordanis Kostelidis 2020-06-17 23:22:38 +0300
  • e3d80f4c66
    Merge pull request #2 from IEEEIHUSerres/add-docker-support Iordanis Kostelidis 2020-06-17 23:21:47 +0300
  • a8ab1de751 adds license and credits for docker Iordanis Kostelidis 2020-06-17 23:19:30 +0300
  • 18dbf0a3fe
    Merge pull request #1 from IEEEIHUSerres/add-docker-support Iordanis Kostelidis 2020-06-17 22:58:24 +0300
  • 33855d7f3e adds requirements for docker on README.md Iordanis Kostelidis 2020-06-17 22:57:20 +0300
  • cd3c88e880 adds guide for docker on README.md Iordanis Kostelidis 2020-06-17 22:54:04 +0300
  • 5aeaa31576 adds newlines end of file Iordanis Kostelidis 2020-06-17 22:37:00 +0300
  • c36052bedb adds docker support for cpu, opencv, openmp and cuda Iordanis Kostelidis 2020-06-17 22:34:32 +0300
  • e08a818d5f Compile fix AlexeyAB 2020-06-15 22:14:01 +0300
  • d724306ee3 New MISH implementation AlexeyAB 2020-06-15 21:55:53 +0300
  • 80fb3cff6d Merge branch 'master' of github.com:AlexeyAB/darknet AlexeyAB 2020-06-15 19:23:11 +0300
  • a553899505 Added Supervised and Unsupervised Contrastive loss for Classifier [net] contrastive=1 unsupervised=1 [contrastive] classes=1000 AlexeyAB 2020-06-15 19:22:47 +0300
  • 040d900350 merge from master Stefano Sinigardi 2020-06-14 23:37:33 +0200
  • 27c49e460d
    Explicitly install CUDA version 10.2 in win-{vcpkg,intlibs}-cuda actions (#5958) Juuso Alasuutari 2020-06-15 00:30:20 +0300
  • 09991d0488 Added Contrastive loss AlexeyAB 2020-06-14 05:11:22 +0300
  • cd98d8da43 Update issue templates Alexey 2020-06-13 18:15:33 +0300
  • 0916c412f6
    Update Readme.md Alexey 2020-06-12 20:30:29 +0300
  • 5691b812b4 Don't free resized image if it's a reference Juuso Alasuutari 2020-06-12 00:17:31 +0300
  • 1f6a6cd71c Give temporary image variable a more descriptive name Juuso Alasuutari 2020-06-12 00:08:49 +0300
  • 0bf989c2cf crop_image() always returns a deep copy, it's safe to call free_image() on it Juuso Alasuutari 2020-06-12 00:05:02 +0300
  • 70f6231d3d Remove some commented-out lines Juuso Alasuutari 2020-06-11 23:59:24 +0300
  • 0a42f8d614
    Update Readme.md Alexey 2020-06-10 17:16:25 +0300
  • 6a6c8319b2
    Update Readme.md Alexey 2020-06-10 17:14:29 +0300
  • 7324fb7b33 fix an openmp related error on macOS tomheaven 2020-06-10 09:05:11 +0800
  • f04b6ad538 fix an openmp related error on macOS tomheaven 2020-06-10 09:00:46 +0800
  • 462980d2d9 Compile fix AlexeyAB 2020-06-08 16:35:24 +0300
  • 7054d0b27e Minor fix AlexeyAB 2020-06-08 16:24:50 +0300
  • 6f4d7a6d1c Added contrastive loss AlexeyAB 2020-06-07 19:10:21 +0300
  • 6c6f04a9b3 [net] mosaic_bound=1 fixed bboxes going out of bounds. Count rewritten bboxes. Changed self-adversarial-training. AlexeyAB 2020-06-07 04:41:01 +0300
  • cd9c0eb763 sample Zhenning Yang 2020-06-06 13:15:40 -0400
  • 918e747c3a updates Zhenning Yang 2020-06-06 13:13:12 -0400
  • cde72f8e0b Fixed [yolo] objectness_smooth=1 and lowest w/h of objectes when mosaic=1 AlexeyAB 2020-06-06 01:47:21 +0300
  • 9a541c1845 update Spain2394 2020-06-04 19:54:40 -0400
  • 47402e980f
    Merge pull request #2 from pjreddie/master Guoping Zhang 2020-06-05 01:32:37 +0800
  • 3708b2e47d
    Update Readme.md Alexey 2020-06-03 16:00:31 +0300
  • 59b4ade9e7
    Update Readme.md Alexey 2020-06-02 14:44:57 +0300
  • 7f1021695a
    Update Readme.md Alexey 2020-06-02 14:38:11 +0300
  • c625cc6503 Added [conv_lstm] time_normalizer=0.5 AlexeyAB 2020-06-01 17:51:29 +0300
  • c5942167d2 Merge branch 'master' of github.com:AlexeyAB/darknet AlexeyAB 2020-06-01 16:47:08 +0300
  • 8aace65afb Fixed mosaic=1 letter_box=1 AlexeyAB 2020-06-01 05:16:45 +0300
  • e455cdef78 Added [conv_lstm] lstm_activation=tanh shortcut=1 AlexeyAB 2020-06-01 05:13:21 +0300
  • e59646002a Update issue templates Alexey 2020-05-31 17:26:28 +0300
  • 16a9b02e64
    Merge pull request #5794 from kay0u/patch-1 Alexey 2020-05-29 21:35:01 +0300
  • b19d883700
    Remove omp in mean_arrays Kayou 2020-05-29 19:08:02 +0200
  • 7e5c3ec24b fix vcpkg freeze Stefano Sinigardi 2020-05-29 18:03:26 +0200
  • 29844ba981
    Merge branch 'master' into dev/cenit/cuda-cmake Stefano Sinigardi 2020-05-29 11:46:47 +0200
  • 269878a44d Added [conv_lstm] bottleneck=1 - to make the layer cheaper AlexeyAB 2020-05-29 03:17:19 +0300
  • 0ba1f84d38 update setup scripts Stefano Sinigardi 2020-05-28 18:56:08 +0200
  • e6469eb071
    Update Readme.md Alexey 2020-05-25 21:29:02 +0300
  • 9697fdb81b Update issue templates Alexey 2020-05-25 16:45:04 +0300
  • 9d3e385fc6 Update issue templates Alexey 2020-05-25 15:43:20 +0300
  • 8d7505f686 Update issue templates Alexey 2020-05-25 15:42:44 +0300
  • 33a3be3344 Update issue templates Alexey 2020-05-25 15:38:13 +0300
  • 8c6e9cde9b Update issue templates Alexey 2020-05-25 15:35:45 +0300
  • c6860f5ef4 Update issue templates Alexey 2020-05-25 15:32:57 +0300
  • 0315ee5896 Update issue templates Alexey 2020-05-25 15:28:23 +0300
  • dcd335f2a3 Update issue templates Alexey 2020-05-25 15:26:27 +0300
  • c0d6b81a78 yolov4.cfg is fixed for stable training without Nan AlexeyAB 2020-05-23 17:52:20 +0300
  • 47013278fa remove linting errors, try to simplify installation even more Stefano Sinigardi 2020-05-23 15:48:30 +0200
  • 6a2e9ca1c6 add compatibility with CMake 3.17 Stefano Sinigardi 2020-05-23 15:48:13 +0200
  • 91e6b8631b add also CUDACXX env variable if missing Stefano Sinigardi 2020-05-23 15:46:13 +0200
  • ba4d8c14c7
    Update Readme.md Alexey 2020-05-22 21:09:53 +0300
  • b44592c25a
    Update Readme.md Alexey 2020-05-22 21:06:36 +0300
  • 4098e4882c Minor fix AlexeyAB 2020-05-20 19:51:06 +0300
  • 03eee3092f Memory leak fix AlexeyAB 2020-05-20 04:22:46 +0300
  • 594f7ce6af Use [net] max_chart_loss=20 by default AlexeyAB 2020-05-18 20:59:15 +0300
  • 103fa8d222 [yolo] max=200 by default for crowds AlexeyAB 2020-05-18 17:00:29 +0300
  • 043711cfcf Fixed ./darknet partial ... for weighted-[shortcut] AlexeyAB 2020-05-18 14:11:15 +0300
  • fcbb74cf98 Added flag averaging detections: ./darknet detector demo ... -avgframes 3 AlexeyAB 2020-05-18 01:31:03 +0300
  • 0ef5052ee5 Fixed crash when train with -map flag AlexeyAB 2020-05-17 00:42:20 +0300
  • f93a73d140
    Do not use vcpkg if not found (#5632) Stefano Sinigardi 2020-05-16 21:01:42 +0200
  • 69d35a6e6b Fixed video input path JaledMC 2020-05-16 17:26:36 +0200
  • 0d764e4ffb
    Merge pull request #5453 from wang-xinyu/master Alexey 2020-05-16 14:57:52 +0300
  • 73768ef6e3 Added [net] max_chart_loss=50 for scale y-axis of Loss-chart AlexeyAB 2020-05-15 18:52:20 +0300
  • bef28445e5 Fixed mish gradient AlexeyAB 2020-05-12 20:15:10 +0300
  • 2500278784 compile fix AlexeyAB 2020-05-12 17:03:26 +0300
  • 0acd6990d8 Minor speedup of Mish/Swish (if train==0 then don't write input of activation to the memory) AlexeyAB 2020-05-12 16:54:13 +0300
  • 83a9cbe5b1 retrigger checks Juan Altmayer Pizzorno 2020-05-11 13:09:01 -0400