OrgLion-ML
/
darknet
mirror of https://github.com/AlexeyAB/darknet.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
2,203 Commits 1 Branch 7 Tags 361 MiB
Commit Graph

2203 Commits

Author SHA1 Message Date
Stefano Sinigardi 108dc8b7fe force ash shell for nuget setup 2021-11-14 13:03:43 +01:00
Stefano Sinigardi 29c78e9e9f
fix unnecessary mono call on windows (#8229) 2021-11-14 11:35:50 +01:00
Stefano Sinigardi 4b35dbbf9a
update to cuda 11.5 (#8174) 2021-11-12 04:58:00 +01:00
AlexeyAB 359001d360 Added yolov4_new.cfg 2021-10-30 06:04:18 +03:00
AlexeyAB 9d40b61975 Added yolov4-csp-s-mish.cfg, yolov4-csp-x-mish.cfg 2021-10-30 05:44:32 +03:00
AlexeyAB 354baa0ce3 Added yolov4-csp-s-mish.cfg, yolov4-csp-x-mish.cfg 2021-10-30 05:43:46 +03:00
Raul Tambre 04b0b15ee4
Link cuda.lib using a preprocessor macro only for non-CMake (#8173) 
Debian CUDA packages don't ship a shim for the Windows SO cuda.lib, so compilation fails.
For CMake it's actually linked using targets, so avoid the preprocessor macro in that case.
 2021-10-24 18:04:17 +02:00
Stefano Sinigardi aef928cc0c
remove travis ci (#8143) 2021-10-08 01:20:20 +03:00
AlexeyAB 568d8c18ab fix for CUDA streams 2021-09-28 23:02:02 +03:00
Sarim Mehdi 9ff8653d99
Made inference faster (this is especially useful when using Yolo9000) (#8009) 
* Update network.c

custom_get_region_detections function now keeps track of class index with the highest probability.

* Update darknet.h

Added best_class_idx to detection struct

* Update darknet.py

added python code for faster negative removal and also faster non-max suppression
 2021-08-25 18:10:21 +03:00
Alexey 2495bd7166
Update README.md 2021-08-22 02:52:15 +03:00
Stefano Sinigardi b8dceb7ed0
[CI] update to cuda11.4 (#7914) 
* [CI] update to cuda11.4

* update setup scripts to cuda 11.4
 2021-07-30 04:05:23 +03:00
AlexeyAB f5007cd0d4 Fixed for-loop var initialization for OpenMP, for Contrastive-loss 2021-07-27 18:43:40 +03:00
Sergey Nuzhny d669680879
Add batch inference on C++ (#7915) 
* Add batch inference on C++

* Return default params

* Add make_nms parameter
 2021-07-18 18:58:01 +03:00
Stéphane Charette 5e73447fa8
allocation and error functions now take a filename, function, and line number to better diagnose where problems happen (#7925) 2021-07-18 06:01:10 +03:00
Stefano Sinigardi cf7b458543
[cmake] update modules (#7891) 
* [cmake] update modules

* [cmake] use proper module name when using vcpkg libs

* apply fixes from review
 2021-07-16 00:30:40 +03:00
8k 4361b7a322
add thresholds (#6831) 2021-07-16 00:28:25 +03:00
Marcus Turewicz 01cf2b4c86
Adds probability to precision recall curve (#7650) 2021-07-16 00:15:47 +03:00
AlexeyAB 9e839c597c darknet_video.py 2021-07-16 00:11:43 +03:00
Yusei-Yagi 7d75fe3c34
Fix bugs in darknet_video.py (#7908) 
* fix VideoWriter frame size

* fix saved image color
 2021-07-16 00:09:42 +03:00
Alexey 1be96802a0
Update README.md 2021-07-12 05:40:11 +03:00
AlexeyAB 2418fa7b08 Added: yolov4-csp-swish.cfg, yolov4-csp-x-swish.cfg, yolov4-csp-x-swish-frozen.cfg 2021-07-12 05:30:51 +03:00
Stefano Sinigardi 9c26b291fa
fix python main module to be compiled with recent build scripts (#7876) 
* fix python main module to be compiled with recent build scripts

* fixes for posix systems
 2021-07-09 14:50:38 +03:00
Alexey 08088dccbb
Update README.md 2021-07-08 01:14:28 +03:00
Alexey 842a602584
Update README.md 2021-07-08 01:13:29 +03:00
AlexeyAB 005513a9db minor fix 2021-07-04 06:38:49 +03:00
AlexeyAB be3985210e fix possible bug 2021-07-04 06:31:45 +03:00
AlexeyAB 335ac662c8 Improved [net] optimized_memory=1 to reduce GPU memory usage. 2021-07-03 04:46:55 +03:00
AlexeyAB b2cb64dffb bug fix 2021-07-03 04:25:42 +03:00
AlexeyAB 49961881ce Merge branch 'master' of https://github.com/AlexeyAB/darknet 2021-07-03 04:18:36 +03:00
AlexeyAB 9c9232d1c3 Bug fix 2021-07-03 04:18:25 +03:00
Alexey 86ced7151a
Update README.md 2021-06-30 06:51:24 +03:00
Alexey 90c15f8f04
Update README.md 2021-06-30 06:37:57 +03:00
AlexeyAB 97ccee5628 minor fix - added: move_window_cv() function 2021-06-29 23:55:34 +03:00
AlexeyAB c2221f07f8 Fixed adversarial training 2021-06-29 23:54:58 +03:00
AlexeyAB 85c6278ef1 Added cfg/yolov4-p5-frozen.cfg for training with pre-trained COCO-weights file yolov4-p5.weights 2021-06-29 22:09:26 +03:00
AlexeyAB 24bb6a8ebf Fixed [<layer>] stopbackward=1, for frozen layers: don't calc batch-norm statistic and don't allocate training-buffers. 2021-06-29 21:32:38 +03:00
AlexeyAB 9a86fce494 Merge branch 'master' of https://github.com/AlexeyAB/darknet 2021-06-27 04:46:16 +03:00
AlexeyAB 28952006b0 Fixed cfg: yolov4-p5.cfg, yolov4-p6.cfg. Added new Weights-files, for: yolov4-csp, yolov4-p5, yolov4-p6, yolov4x-mish 2021-06-27 04:46:03 +03:00
Stefano Sinigardi 8bbdb6b77c
[CI] test vcpkg nightly builds (#7826) 2021-06-24 10:23:45 +02:00
AlexeyAB 795474f0ed Merge branch 'master' of https://github.com/AlexeyAB/darknet 2021-06-22 00:31:53 +03:00
AlexeyAB 46dea82bce Added yolov4-p5.cfg and yolov4-p6.cfg 2021-06-22 00:31:38 +03:00
Alexey 070ed01e26
Update README.md 2021-06-21 23:47:53 +03:00
Stefano Sinigardi 3fafd211b9
[autorebase] bump to v1.5 (#7819) 2021-06-21 11:07:31 +02:00
Stefano Sinigardi 6c171a4cea
add automatic rebase github action (#7809) 
* add automatic rebase action

* do not upload artifacts from PRs

* use proper symbol in build.ps1

* fix lib naming

* fix folder for dll artifacts
 2021-06-18 01:15:55 +03:00
Double 396b5299a7
fix: flush buffer to weights-file when writing each layer (#7747) 
Co-authored-by: Double.c <double.c@inwinstack.com>
 2021-06-16 19:25:00 +03:00
Alexey 00b8e29b68
Update README.md 2021-06-12 02:53:08 +03:00
AlexeyAB 108c8cfd80 Merge branch 'master' of https://github.com/AlexeyAB/darknet 2021-06-11 03:38:35 +03:00
AlexeyAB 67543e27aa fix darknet_no_gpu.vcxproj 2021-06-11 03:38:20 +03:00
Ievgen Popovych 894aada70b
region_layer: Populate out_h/w/c fileds for consitency (#7783) 2021-06-11 03:36:12 +03:00