|
|
b1b2d5cfef
|
Update Readme.md
|
2020-04-24 04:02:02 +03:00 |
|
|
|
3d4242a6e5
|
Update Readme.md
|
2020-04-23 15:23:39 +03:00 |
|
|
|
4131ae1d6e
|
Yolo v4
|
2020-04-23 14:49:06 +03:00 |
|
|
|
b573eab63f
|
Minor fix
|
2020-04-22 20:49:27 +03:00 |
|
|
|
88f28f7fcc
|
Merge pull request #4099 from enesozi/master
Batch inference added
|
2020-04-22 14:46:38 +03:00 |
|
|
|
7c1f9105f0
|
Minor fix
|
2020-04-21 04:25:43 +03:00 |
|
|
|
1fb4606d5d
|
Minor fix
|
2020-04-19 19:45:04 +03:00 |
|
|
|
4786d557f9
|
Fixed memory leak for Training with OPENCV=1 - create pool of CPU-threads only once (bug in OpenCV - it allocates memory for each thread without deallocation)
|
2020-04-18 19:01:20 +03:00 |
|
|
|
66c809fe78
|
Compile fix
|
2020-04-18 16:55:53 +03:00 |
|
|
|
56a7ccf07e
|
Improved FPS and and memory leak for ./darknet detector demo ... (bug in OpenCV - allocates memory for each thread without deallocation)
|
2020-04-18 16:40:22 +03:00 |
|
|
|
5dc7d8074d
|
Tensor Cores fixed
|
2020-04-18 03:25:16 +03:00 |
|
|
|
d4b2ed9d22
|
Added param train_only_bn=1 to train only Batch-normalization for this and previous layers
|
2020-04-17 19:06:15 +03:00 |
|
|
|
342a8d1561
|
Fixed std thread
|
2020-04-13 23:03:50 +03:00 |
|
|
|
00193efd83
|
Use C++ thread for demo.c
|
2020-04-13 22:52:20 +03:00 |
|
|
|
9e932d65c5
|
Fixed compilation for ZED camera
|
2020-04-11 15:16:40 +03:00 |
|
|
|
08b1c20e26
|
Fixed seen for multu-GPU training
|
2020-04-11 14:57:42 +03:00 |
|
|
|
3d9aa2af47
|
./darknet detector calc_anchors calculates both anchors= and counters_per_class=
|
2020-04-07 22:40:20 +03:00 |
|
|
|
a9bae4f032
|
Fixed DropBlock
|
2020-04-03 20:33:36 +03:00 |
|
|
|
afb4cc4766
|
Fix for iou_thresh_kind= param
|
2020-04-02 19:04:08 +03:00 |
|
|
|
2f9f4a40d3
|
Minor fix
|
2020-04-01 15:02:21 +03:00 |
|
|
|
68062fa4fb
|
Added param to [yolo] iou_thresh_kind=giou iou_thresh=0.213
|
2020-03-31 21:12:41 +03:00 |
|
|
|
cedfd792c6
|
Merge pull request #5145 from keita-honsho/fix_errors
fix some errors.
|
2020-03-31 20:55:26 +03:00 |
|
|
|
4cd21785fc
|
Minor fix
|
2020-03-31 19:15:33 +03:00 |
|
|
|
07e4a7d7b6
|
fix error quantize_image's result type.
fix warning printf params, in backward_network_gpu.
|
2020-03-31 11:27:48 +09:00 |
|
|
|
0e06337150
|
minor fix
|
2020-03-29 20:03:33 +03:00 |
|
|
|
6a77b8dc9a
|
Show receptive field
|
2020-03-29 20:00:03 +03:00 |
|
|
|
9a2344759b
|
adversarial training learning rate fix
|
2020-03-27 20:08:40 +03:00 |
|
|
|
4f62a01cd3
|
Fixed adversarial attack and training
|
2020-03-27 16:29:41 +03:00 |
|
|
|
2614a231f0
|
Fixed relu to lrelu in fuse_conv_batchnorm()
|
2020-03-26 15:37:18 +03:00 |
|
|
|
d6181c67cd
|
Fix one relu() to lrelu() for weighted-shortcut
|
2020-03-26 15:28:20 +03:00 |
|
|
|
8cb1f0c44f
|
Minor fix
|
2020-03-26 02:34:01 +03:00 |
|
|
|
9919612006
|
Added attention during training: [net] adversarial_lr=0.05 attention=1
|
2020-03-26 01:25:04 +03:00 |
|
|
|
978c2e00ed
|
self-adversarial training
|
2020-03-26 00:16:57 +03:00 |
|
|
|
39d79785ba
|
Minor fix
|
2020-03-25 03:33:26 +03:00 |
|
|
|
b3c9f373d1
|
adversarial attack - minor improvement
|
2020-03-25 02:35:05 +03:00 |
|
|
|
aa4f9975a7
|
adversarial attack - fixed
|
2020-03-24 23:02:44 +03:00 |
|
|
|
fdb61f8cb4
|
Minor fix
|
2020-03-24 22:30:30 +03:00 |
|
|
|
5e904d5b1e
|
adversarial attack - experimental
|
2020-03-24 22:23:15 +03:00 |
|
|
|
a234a50223
|
Added efficientnet-lite3.cfg and activation=relu6
|
2020-03-23 21:56:32 +03:00 |
|
|
|
bb5eb481d1
|
Minor fix
|
2020-03-23 20:03:56 +03:00 |
|
|
|
864d1062f8
|
Minor fix
|
2020-03-23 00:29:37 +03:00 |
|
|
|
5f5ec84e40
|
Merge pull request #4783 from kseniaryabinova/chart_argument
Added -chart argument
|
2020-03-23 00:19:29 +03:00 |
|
|
|
17dd701618
|
Merge pull request #4976 from mmaaz60/monitor_training
Monitor training
|
2020-03-23 00:15:57 +03:00 |
|
|
|
9407a81c9d
|
Fix: Correct calculation of training time remaininig when resume training (src/classifier.c)
Output training time remaining to console if darknet is not compiled with OpenCV
|
2020-03-22 20:49:21 +05:00 |
|
|
|
8800212faa
|
Adds support to print remaining training time on console if darknet is not compiled with OpenCV
|
2020-03-22 20:40:26 +05:00 |
|
|
|
bd7690b019
|
Fix: Correct remaining training time in case someone resume training (src/detector.c)
|
2020-03-22 20:38:47 +05:00 |
|
|
|
5c2ddd301e
|
Added param to cfg: [net] gaussian_noise=40
|
2020-03-22 17:38:07 +03:00 |
|
|
|
570f595247
|
Remove redundant code from classifier.c
|
2020-03-20 12:42:59 +05:00 |
|
|
|
1a6dd593c5
|
Fix: Use iteration instead of i for current iteration.
Simplified the time remaining calculation logic.
|
2020-03-19 17:15:42 +05:00 |
|
|
|
02e7109b8b
|
Merge branch 'master' into monitor_training
|
2020-03-19 13:31:51 +05:00 |
|
