|
|
0fa9c8f105
|
Added groups= and groupd_id= params to the [route] layer
|
2019-11-08 01:06:49 +03:00 |
|
|
|
704bd1138e
|
Fixed params.net in parse_convolutional() and parse_route()
|
2019-11-07 23:49:10 +03:00 |
|
|
|
c7c7078de7
|
Added URL to the description of activation MISH on GitHub
|
2019-11-07 01:18:47 +03:00 |
|
|
|
bf8ea4183d
|
Added MISH activation, use activation=mish in [convolutional] layers
|
2019-11-07 01:15:58 +03:00 |
|
|
|
d628e8eab7
|
Fixed darknet.py for Uncertainty (gaussian_yolo_layer)
|
2019-11-05 16:57:03 +03:00 |
|
|
|
e345b8793c
|
Fixed ./darknet partial for Anti-aliasing
|
2019-11-05 16:06:24 +03:00 |
|
|
|
7018be435f
|
Compile fix
|
2019-11-04 14:45:34 +03:00 |
|
|
|
52e3bb252f
|
Use non-blocking sockets for JSON_server and MJPEG_server
|
2019-11-04 14:39:20 +03:00 |
|
|
|
4d9addedd7
|
Fixed legacy compilation using MSVS (gaussian_yolo_layer)
|
2019-10-29 14:55:12 +03:00 |
|
|
|
6e73633925
|
Fixed Blur and Try to use new Assisted Excitation.
|
2019-10-29 14:53:57 +03:00 |
|
|
|
69b7a19f5b
|
minor fix
|
2019-10-27 02:45:31 +03:00 |
|
|
|
f8c72acd42
|
Added scale_x_y to [Gaussian_yolo]. Fixed blur=10;
|
2019-10-26 14:47:00 +03:00 |
|
|
|
fa7687e6b5
|
cfg-param blur > 1 is fixed (bilateral filter removes textures)
|
2019-10-26 01:40:59 +03:00 |
|
|
|
29c71a190a
|
Fixed antialiasing=1 for [convolutional] layer
|
2019-10-26 01:30:11 +03:00 |
|
|
|
f18338de26
|
Fixed [Gaussian_yolo] layer (tested for training and detection)
|
2019-10-26 01:29:41 +03:00 |
|
|
|
72f6de30b2
|
another compile fix
|
2019-10-25 22:14:27 +03:00 |
|
|
|
24788b8061
|
Compile fix
|
2019-10-25 20:56:58 +03:00 |
|
|
|
b3a2495298
|
Added Gaussian YOLOv3 layer [Gaussian_yolo]
|
2019-10-25 20:47:17 +03:00 |
|
|
|
bb7d69941c
|
Added debugging info for Training Classifier for case: Too many or too few labels
|
2019-10-21 15:49:46 +03:00 |
|
|
|
e6486ab594
|
Use ignore_thresh only if class_id matched. Temporary changed Assisted_Excitation (reduces background activations rather than enhancing objects activations). Added antialiasiong=2 for 2x2.
|
2019-10-21 15:33:01 +03:00 |
|
|
|
2eb68d5177
|
Update Readme.md
|
2019-10-19 15:09:15 +03:00 |
|
|
|
5dfbbfdb5f
|
Merge pull request #3375 from acxz/build_instructions
Build instructions
|
2019-10-19 15:05:20 +03:00 |
|
|
|
0823d04247
|
Add readability changes
Make CMake-GUI install more visible than vcpkg install
|
2019-10-18 19:24:49 -04:00 |
|
|
|
53160fa666
|
Merge pull request #3501 from shooorf/master
Check if image does not require resizing
|
2019-10-08 17:07:12 +03:00 |
|
|
|
e24c96dc8b
|
Update Readme.md
|
2019-10-03 17:28:06 +03:00 |
|
|
|
7181c7435f
|
Update readme.md
|
2019-10-03 16:40:08 +03:00 |
|
|
|
ff0b739e43
|
Update Readme.md
|
2019-10-03 12:12:00 +03:00 |
|
|
|
2fa539779f
|
Readme.md
|
2019-09-27 22:35:56 +03:00 |
|
|
|
05545b2607
|
Create FUNDING.yml
|
2019-09-24 15:31:28 +03:00 |
|
|
|
6f4d93bb9f
|
Update readme.md
|
2019-09-22 17:38:42 +03:00 |
|
|
|
b918bf0329
|
Merge pull request #3946 from keko950/batch
Removed useless snippet that only breaks if batch processing is enabled
|
2019-09-18 15:36:58 +03:00 |
|
|
|
fa74f691cb
|
Removed useless snippet that only breaks if batch processing is enabled
|
2019-09-18 13:49:54 +02:00 |
|
|
|
1c71f00153
|
Fixed shortcut_layer for antialiasing
|
2019-09-07 15:15:19 +03:00 |
|
|
|
b0b1584a06
|
Minor fix
|
2019-09-06 01:53:36 +03:00 |
|
|
|
e33019e669
|
Fixed stride_x and stride_y. Fixed AntiAliasing. Minor fixes.
|
2019-09-05 14:47:26 +03:00 |
|
|
|
f6fa4a56d9
|
compile fix
|
2019-09-04 19:58:36 +03:00 |
|
|
|
1e0b50c29e
|
Added yolov3-tiny-prn.cfg and enet-coco.cfg (EfficientNetb0-Yolo- 45.5% mAP@0.5 - 3.7 BFlops) https://github.com/WongKinYiu/PartialResidualNetworks
|
2019-09-04 18:51:40 +03:00 |
|
|
|
9c02df864e
|
Fixed assisted_excitation and added also for [shortcut] layer
|
2019-09-04 18:50:56 +03:00 |
|
|
|
be5d0d6693
|
Added assisted_excitation=1 for [convolutional] layer on GPU
|
2019-09-03 01:35:05 +03:00 |
|
|
|
a63782ca89
|
Added: efficientnet_b0.cfg
|
2019-09-02 15:55:05 +03:00 |
|
|
|
9e26472b1a
|
Added antialiasing=1 param for [maxpool]-layer on GPU and CPU
|
2019-09-02 15:25:42 +03:00 |
|
|
|
80ceee4fca
|
Added antialiasing=1 param for [convolutional]-layer on CPU (only forward inference)
|
2019-09-02 14:24:54 +03:00 |
|
|
|
11b8b2bddf
|
Added antialiasing=1 param for [convolutional]-layer on GPU
|
2019-09-02 02:13:11 +03:00 |
|
|
|
a4a06c337a
|
Addex stride_x= and stride_y= for [convolutional] layer for MatrixNet
|
2019-09-01 16:07:33 +03:00 |
|
|
|
102ab710a9
|
Fixed fuse_conv_batchnorm() for TridentNet
|
2019-09-01 14:11:17 +03:00 |
|
|
|
2a382e5a4b
|
Fix training TridentNet
|
2019-08-30 20:28:56 +03:00 |
|
|
|
660a9b225c
|
compile fix
|
2019-08-30 00:29:26 +03:00 |
|
|
|
58906ef812
|
minor fix: Unexpected end of wights-file!
|
2019-08-29 23:34:04 +03:00 |
|
|
|
4acf924aaf
|
minor fix
|
2019-08-29 18:30:33 +03:00 |
|
|
|
8d80a65288
|
Max pool layer can use stride=2 or stride_x=2 stride_y=4 (isn't tested well)
|
2019-08-27 14:04:39 +03:00 |
|
