OrgLion-ML
/
darknet
mirror of https://github.com/AlexeyAB/darknet.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
998 Commits 1 Branch 7 Tags 361 MiB
C 62%
Cuda 14.6%
C++ 12.5%
Python 4.5%
PowerShell 2.9%
Other 3.5%
Go to file
AlexeyAB 29f509e2b7 Added yolov3-tiny_3l.cfg for small objects 2019-01-04 18:46:10 +03:00
.circleci Added CUDNN_HALF to the Makefile 2018-04-16 17:52:03 +03:00
3rdparty To Build on Windows - added: darknet.sln, darknet.vcxproj, pthreadVC2.dll, pthreadGC2.dll 2016-12-02 17:29:29 +03:00
build/darknet Added yolov3-tiny_3l.cfg for small objects 2019-01-04 18:46:10 +03:00
cfg Added yolov3-tiny_3l.cfg for small objects 2019-01-04 18:46:10 +03:00
data Added yolov3-openimages.cfg and SELU activation 2018-10-15 20:16:08 +03:00
scripts get_coco_dataset.sh downloads MS COCO dataset from images.cocodataset.org 2018-12-21 23:07:37 +03:00
src Minor fix 2018-12-29 22:56:38 +03:00
.gitignore fix build error, add c# wrapper class 2018-06-24 22:45:07 +02:00
LICENSE Update LICENSE 2016-07-29 13:11:41 -07:00
Makefile Compile fix 2018-12-07 22:52:07 +03:00
README.md Update Readme.md 2018-12-23 01:00:16 +03:00
darknet.py fix wrong path 2018-10-24 12:47:19 +07:00
image_yolov2.sh Added Yolo v2 bash files 2018-03-28 23:33:03 +03:00
image_yolov3.sh Added Yolo v3 2018-03-28 02:59:03 +03:00
mjpeg_stream.sh Changed cmd and sh files. Added new cfg files for old Yolo v2. 2018-03-28 23:23:37 +03:00
net_cam_v3.sh Fixed typo in weights file 2018-12-26 17:21:54 +04:00
video_v2.sh Added Yolo v2 bash files 2018-03-28 23:33:03 +03:00
video_yolov3.sh Added Yolo v3 2018-03-28 02:59:03 +03:00

README.md

Yolo-v3 and Yolo-v2 for Windows and Linux

(neural network for object detection) - Tensor Cores can be used on Linux and Windows

CircleCI

  1. Improvements in this repository
  2. How to use
  3. How to compile on Linux
  4. How to compile on Windows
  5. How to train (Pascal VOC Data)
  6. How to train (to detect your custom objects)
  7. When should I stop training
  8. How to calculate mAP on PascalVOC 2007
  9. How to improve object detection
  10. How to mark bounded boxes of objects and create annotation files
  11. Using Yolo9000
  12. How to use Yolo as DLL
Darknet Logo   map_fps mAP (AP50) https://pjreddie.com/media/files/papers/YOLOv3.pdf

"You Only Look Once: Unified, Real-Time Object Detection (versions 2 & 3)"

A Yolo cross-platform Windows and Linux version (for object detection). Contributtors: https://github.com/pjreddie/darknet/graphs/contributors

This repository is forked from Linux-version: https://github.com/pjreddie/darknet

More details: http://pjreddie.com/darknet/yolo/

This repository supports:

  • both Windows and Linux
  • both OpenCV 2.x.x and OpenCV <= 3.4.0 (3.4.1 and higher isn't supported)
  • both cuDNN v5-v7
  • CUDA >= 7.5
  • also create SO-library on Linux and DLL-library on Windows
Requires:
Pre-trained models for different cfg-files can be downloaded from (smaller -> faster & lower quality):

Put it near compiled: darknet.exe

You can get cfg-files by path: darknet/cfg/

Examples of results:

Everything Is AWESOME

Others: https://www.youtube.com/channel/UC7ev3hNVkx4DzZ3LO19oebg

Improvements in this repository

  • added support for Windows
  • improved binary neural network performance 2x-4x times for Detection on CPU and GPU if you trained your own weights by using this XNOR-net model (bit-1 inference) : https://github.com/AlexeyAB/darknet/blob/master/cfg/yolov3-tiny_xnor.cfg
  • improved neural network performance ~7% by fusing 2 layers into 1: Convolutional + Batch-norm
  • improved neural network performance Detection 3x times, Training 2 x times on GPU Volta (Tesla V100, Titan V, ...) using Tensor Cores if CUDNN_HALF defined in the Makefile or darknet.sln
  • improved performance ~1.2x times on FullHD, ~2x times on 4K, for detection on the video (file/stream) using darknet detector demo...
  • improved performance 3.5 X times of data augmentation for training (using OpenCV SSE/AVX functions instead of hand-written functions) - removes bottleneck for training on multi-GPU or GPU Volta
  • improved performance of detection and training on Intel CPU with AVX (Yolo v3 ~85%, Yolo v2 ~10%)
  • fixed usage of [reorg]-layer
  • optimized memory allocation during network resizing when random=1
  • optimized initialization GPU for detection - we use batch=1 initially instead of re-init with batch=1
  • added correct calculation of mAP, F1, IoU, Precision-Recall using command darknet detector map...
  • added drawing of chart of average loss during training
  • added calculation of anchors for training
  • added example of Detection and Tracking objects: https://github.com/AlexeyAB/darknet/blob/master/src/yolo_console_dll.cpp
  • fixed code for use Web-cam on OpenCV 3.x
  • run-time tips and warnings if you use incorrect cfg-file or dataset
  • many other fixes of code...

And added manual - How to train Yolo v3/v2 (to detect your custom objects)

Also, you might be interested in using a simplified repository where is implemented INT8-quantization (+30% speedup and -1% mAP reduced): https://github.com/AlexeyAB/yolo2_light

How to use:

Example of usage in cmd-files from build\darknet\x64\:

  • darknet_yolo_v3.cmd - initialization with 236 MB Yolo v3 COCO-model yolov3.weights & yolov3.cfg and show detection on the image: dog.jpg

  • darknet_voc.cmd - initialization with 194 MB VOC-model yolo-voc.weights & yolo-voc.cfg and waiting for entering the name of the image file

  • darknet_demo_voc.cmd - initialization with 194 MB VOC-model yolo-voc.weights & yolo-voc.cfg and play your video file which you must rename to: test.mp4

  • darknet_demo_store.cmd - initialization with 194 MB VOC-model yolo-voc.weights & yolo-voc.cfg and play your video file which you must rename to: test.mp4, and store result to: res.avi

  • darknet_net_cam_voc.cmd - initialization with 194 MB VOC-model, play video from network video-camera mjpeg-stream (also from you phone)

  • darknet_web_cam_voc.cmd - initialization with 194 MB VOC-model, play video from Web-Camera number #0

  • darknet_coco_9000.cmd - initialization with 186 MB Yolo9000 COCO-model, and show detection on the image: dog.jpg

  • darknet_coco_9000_demo.cmd - initialization with 186 MB Yolo9000 COCO-model, and show detection on the video (if it is present): street4k.mp4, and store result to: res.avi

How to use on the command line:

On Linux use ./darknet instead of darknet.exe, like this:./darknet detector test ./cfg/coco.data ./cfg/yolov3.cfg ./yolov3.weights

  • Yolo v3 COCO - image: darknet.exe detector test data/coco.data cfg/yolov3.cfg yolov3.weights -i 0 -thresh 0.25
  • Output coordinates of objects: darknet.exe detector test data/coco.data yolov3.cfg yolov3.weights -ext_output dog.jpg
  • Yolo v3 COCO - video: darknet.exe detector demo data/coco.data cfg/yolov3.cfg yolov3.weights -ext_output test.mp4
  • Yolo v3 COCO - WebCam 0: darknet.exe detector demo data/coco.data cfg/yolov3.cfg yolov3.weights -c 0
  • Yolo v3 COCO for net-videocam - Smart WebCam: darknet.exe detector demo data/coco.data cfg/yolov3.cfg yolov3.weights http://192.168.0.80:8080/video?dummy=param.mjpg
  • Yolo v3 - save result to the file res.avi: darknet.exe detector demo data/coco.data cfg/yolov3.cfg yolov3.weights -thresh 0.25 test.mp4 -out_filename res.avi
  • Yolo v3 Tiny COCO - video: darknet.exe detector demo data/coco.data cfg/yolov3-tiny.cfg yolov3-tiny.weights test.mp4
  • Yolo v3 Tiny on GPU #0: darknet.exe detector demo data/coco.data cfg/yolov3-tiny.cfg yolov3-tiny.weights -i 0 test.mp4
  • Alternative method Yolo v3 COCO - image: darknet.exe detect cfg/yolov3.cfg yolov3.weights -i 0 -thresh 0.25
  • 186 MB Yolo9000 - image: darknet.exe detector test cfg/combine9k.data yolo9000.cfg yolo9000.weights
  • Remeber to put data/9k.tree and data/coco9k.map under the same folder of your app if you use the cpp api to build an app
  • To process a list of images data/train.txt and save results of detection to result.txt use:
    darknet.exe detector test cfg/coco.data yolov3.cfg yolov3.weights -dont_show -ext_output < data/train.txt > result.txt
For using network video-camera mjpeg-stream with any Android smartphone:

  1. Download for Android phone mjpeg-stream soft: IP Webcam / Smart WebCam

  2. Connect your Android phone to computer by WiFi (through a WiFi-router) or USB

  3. Start Smart WebCam on your phone

  4. Replace the address below, on shown in the phone application (Smart WebCam) and launch:

  • Yolo v3 COCO-model: darknet.exe detector demo data/coco.data yolov3.cfg yolov3.weights http://192.168.0.80:8080/video?dummy=param.mjpg -i 0

How to compile on Linux:

Just do make in the darknet directory. Before make, you can set such options in the Makefile: link

  • GPU=1 to build with CUDA to accelerate by using GPU (CUDA should be in /usr/local/cuda)
  • CUDNN=1 to build with cuDNN v5-v7 to accelerate training by using GPU (cuDNN should be in /usr/local/cudnn)
  • CUDNN_HALF=1 to build for Tensor Cores (on Titan V / Tesla V100 / DGX-2 and later) speedup Detection 3x, Training 2x
  • OPENCV=1 to build with OpenCV 3.x/2.4.x - allows to detect on video files and video streams from network cameras or web-cams
  • DEBUG=1 to bould debug version of Yolo
  • OPENMP=1 to build with OpenMP support to accelerate Yolo by using multi-core CPU
  • LIBSO=1 to build a library darknet.so and binary runable file uselib that uses this library. Or you can try to run so LD_LIBRARY_PATH=./:$LD_LIBRARY_PATH ./uselib test.mp4 How to use this SO-library from your own code - you can look at C++ example: https://github.com/AlexeyAB/darknet/blob/master/src/yolo_console_dll.cpp or use in such a way: LD_LIBRARY_PATH=./:$LD_LIBRARY_PATH ./uselib data/coco.names cfg/yolov3.cfg yolov3.weights test.mp4

To run Darknet on Linux use examples from this article, just use ./darknet instead of darknet.exe, i.e. use this command: ./darknet detector test ./cfg/coco.data ./cfg/yolov3.cfg ./yolov3.weights

How to compile on Windows:

  1. If you have MSVS 2015, CUDA 10.0, cuDNN 7.4 and OpenCV 3.x (with paths: C:\opencv_3.0\opencv\build\include & C:\opencv_3.0\opencv\build\x64\vc14\lib), then start MSVS, open build\darknet\darknet.sln, set x64 and Release https://hsto.org/webt/uh/fk/-e/uhfk-eb0q-hwd9hsxhrikbokd6u.jpeg and do the: Build -> Build darknet. Also add Windows system variable cudnn with path to CUDNN: https://hsto.org/files/a49/3dc/fc4/a493dcfc4bd34a1295fd15e0e2e01f26.jpg NOTE: If installing OpenCV, use OpenCV 3.4.0 or earlier. This is a bug in OpenCV 3.4.1 in the C API (see #500).

    1.1. Find files opencv_world320.dll and opencv_ffmpeg320_64.dll (or opencv_world340.dll and opencv_ffmpeg340_64.dll) in C:\opencv_3.0\opencv\build\x64\vc14\bin and put it near with darknet.exe

    1.2 Check that there are bin and include folders in the C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1 if aren't, then copy them to this folder from the path where is CUDA installed

    1.3. To install CUDNN (speedup neural network), do the following:

    1.4. If you want to build without CUDNN then: open \darknet.sln -> (right click on project) -> properties -> C/C++ -> Preprocessor -> Preprocessor Definitions, and remove this: CUDNN;

  2. If you have other version of CUDA (not 10.0) then open build\darknet\darknet.vcxproj by using Notepad, find 2 places with "CUDA 10.0" and change it to your CUDA-version, then do step 1

  3. If you don't have GPU, but have MSVS 2015 and OpenCV 3.0 (with paths: C:\opencv_3.0\opencv\build\include & C:\opencv_3.0\opencv\build\x64\vc14\lib), then start MSVS, open build\darknet\darknet_no_gpu.sln, set x64 and Release, and do the: Build -> Build darknet_no_gpu

  4. If you have OpenCV 2.4.13 instead of 3.0 then you should change pathes after \darknet.sln is opened

    4.1 (right click on project) -> properties -> C/C++ -> General -> Additional Include Directories: C:\opencv_2.4.13\opencv\build\include

    4.2 (right click on project) -> properties -> Linker -> General -> Additional Library Directories: C:\opencv_2.4.13\opencv\build\x64\vc14\lib

  5. If you have GPU with Tensor Cores (nVidia Titan V / Tesla V100 / DGX-2 and later) speedup Detection 3x, Training 2x: \darknet.sln -> (right click on project) -> properties -> C/C++ -> Preprocessor -> Preprocessor Definitions, and add here: CUDNN_HALF;

    Note: CUDA must be installed only after that MSVS2015 had been installed.

How to compile (custom):

Also, you can to create your own darknet.sln & darknet.vcxproj, this example for CUDA 9.1 and OpenCV 3.0

Then add to your created project:

  • (right click on project) -> properties -> C/C++ -> General -> Additional Include Directories, put here:

C:\opencv_3.0\opencv\build\include;..\..\3rdparty\include;%(AdditionalIncludeDirectories);$(CudaToolkitIncludeDir);$(cudnn)\include

C:\opencv_3.0\opencv\build\x64\vc14\lib;$(CUDA_PATH)lib\$(PlatformName);$(cudnn)\lib\x64;%(AdditionalLibraryDirectories)

  • (right click on project) -> properties -> Linker -> Input -> Additional dependecies, put here:

..\..\3rdparty\lib\x64\pthreadVC2.lib;cublas.lib;curand.lib;cudart.lib;cudnn.lib;%(AdditionalDependencies)

  • (right click on project) -> properties -> C/C++ -> Preprocessor -> Preprocessor Definitions

OPENCV;_TIMESPEC_DEFINED;_CRT_SECURE_NO_WARNINGS;_CRT_RAND_S;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions)

  • compile to .exe (X64 & Release) and put .dll-s near with .exe: https://hsto.org/webt/uh/fk/-e/uhfk-eb0q-hwd9hsxhrikbokd6u.jpeg

    • pthreadVC2.dll, pthreadGC2.dll from \3rdparty\dll\x64

    • cusolver64_91.dll, curand64_91.dll, cudart64_91.dll, cublas64_91.dll - 91 for CUDA 9.1 or your version, from C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\bin

    • For OpenCV 3.2: opencv_world320.dll and opencv_ffmpeg320_64.dll from C:\opencv_3.0\opencv\build\x64\vc14\bin

    • For OpenCV 2.4.13: opencv_core2413.dll, opencv_highgui2413.dll and opencv_ffmpeg2413_64.dll from C:\opencv_2.4.13\opencv\build\x64\vc14\bin

How to train (Pascal VOC Data):

  1. Download pre-trained weights for the convolutional layers (154 MB): http://pjreddie.com/media/files/darknet53.conv.74 and put to the directory build\darknet\x64

  2. Download The Pascal VOC Data and unpack it to directory build\darknet\x64\data\voc will be created dir build\darknet\x64\data\voc\VOCdevkit\:

    2.1 Download file voc_label.py to dir build\darknet\x64\data\voc: http://pjreddie.com/media/files/voc_label.py

  3. Download and install Python for Windows: https://www.python.org/ftp/python/3.5.2/python-3.5.2-amd64.exe

  4. Run command: python build\darknet\x64\data\voc\voc_label.py (to generate files: 2007_test.txt, 2007_train.txt, 2007_val.txt, 2012_train.txt, 2012_val.txt)

  5. Run command: type 2007_train.txt 2007_val.txt 2012_*.txt > train.txt

  6. Set batch=64 and subdivisions=8 in the file yolov3-voc.cfg: link

  7. Start training by using train_voc.cmd or by using the command line:

    darknet.exe detector train data/voc.data cfg/yolov3-voc.cfg darknet53.conv.74

(Note: To disable Loss-Window use flag -dont_show. If you are using CPU, try darknet_no_gpu.exe instead of darknet.exe.)

If required change pathes in the file build\darknet\x64\data\voc.data

More information about training by the link: http://pjreddie.com/darknet/yolo/#train-voc

Note: If during training you see nan values for avg (loss) field - then training goes wrong, but if nan is in some other lines - then training goes well.

How to train with multi-GPU:

  1. Train it first on 1 GPU for like 1000 iterations: darknet.exe detector train data/voc.data cfg/yolov3-voc.cfg darknet53.conv.74

  2. Then stop and by using partially-trained model /backup/yolov3-voc_1000.weights run training with multigpu (up to 4 GPUs): darknet.exe detector train data/voc.data cfg/yolov3-voc.cfg /backup/yolov3-voc_1000.weights -gpus 0,1,2,3

Only for small datasets sometimes better to decrease learning rate, for 4 GPUs set learning_rate = 0.00025 (i.e. learning_rate = 0.001 / GPUs). In this case also increase 4x times burn_in = and max_batches = in your cfg-file. I.e. use burn_in = 4000 instead of 1000.

https://groups.google.com/d/msg/darknet/NbJqonJBTSY/Te5PfIpuCAAJ

How to train (to detect your custom objects):

(to train old Yolo v2 yolov2-voc.cfg, yolov2-tiny-voc.cfg, yolo-voc.cfg, yolo-voc.2.0.cfg, ... click by the link)

Training Yolo v3:

  1. Create file yolo-obj.cfg with the same content as in yolov3.cfg (or copy yolov3.cfg to yolo-obj.cfg) and:

So if classes=1 then should be filters=18. If classes=2 then write filters=21.

(Do not write in the cfg-file: filters=(classes + 5)x3)

(Generally filters depends on the classes, coords and number of masks, i.e. filters=(classes + coords + 1)*<number of mask>, where mask is indices of anchors. If mask is absence, then filters=(classes + coords + 1)*num)

So for example, for 2 objects, your file yolo-obj.cfg should differ from yolov3.cfg in such lines in each of 3 [yolo]-layers:

[convolutional]
filters=21

[region]
classes=2

  1. Create file obj.names in the directory build\darknet\x64\data\, with objects names - each in new line

  2. Create file obj.data in the directory build\darknet\x64\data\, containing (where classes = number of objects):

classes= 2
train  = data/train.txt
valid  = data/test.txt
names = data/obj.names
backup = backup/

  1. Put image-files (.jpg) of your objects in the directory build\darknet\x64\data\obj\

  2. You should label each object on images from your dataset. Use this visual GUI-software for marking bounded boxes of objects and generating annotation files for Yolo v2 & v3: https://github.com/AlexeyAB/Yolo_mark

It will create .txt-file for each .jpg-image-file - in the same directory and with the same name, but with .txt-extension, and put to file: object number and object coordinates on this image, for each object in new line: <object-class> <x> <y> <width> <height>

Where:

  • <object-class> - integer object number from 0 to (classes-1)
  • <x_center> <y_center> <width> <height> - float values relative to width and height of image, it can be equal from (0.0 to 1.0]
  • for example: <x> = <absolute_x> / <image_width> or <height> = <absolute_height> / <image_height>
  • atention: <x_center> <y_center> - are center of rectangle (are not top-left corner)

For example for img1.jpg you will be created img1.txt containing:

1 0.716797 0.395833 0.216406 0.147222
0 0.687109 0.379167 0.255469 0.158333
1 0.420312 0.395833 0.140625 0.166667
  1. Create file train.txt in directory build\darknet\x64\data\, with filenames of your images, each filename in new line, with path relative to darknet.exe, for example containing:
data/obj/img1.jpg
data/obj/img2.jpg
data/obj/img3.jpg

  1. Download pre-trained weights for the convolutional layers (154 MB): https://pjreddie.com/media/files/darknet53.conv.74 and put to the directory build\darknet\x64

  2. Start training by using the command line: darknet.exe detector train data/obj.data yolo-obj.cfg darknet53.conv.74

    To train on Linux use command: ./darknet detector train data/obj.data yolo-obj.cfg darknet53.conv.74 (just use ./darknet instead of darknet.exe)

    • (file yolo-obj_last.weights will be saved to the build\darknet\x64\backup\ for each 100 iterations)
    • (file yolo-obj_xxxx.weights will be saved to the build\darknet\x64\backup\ for each 1000 iterations)
    • (To disable Loss-Window use darknet.exe detector train data/obj.data yolo-obj.cfg darknet53.conv.74 -dont_show, if you train on computer without monitor like a cloud Amazaon EC2)

8.1. For training with mAP (mean average precisions) calculation for each 4 Epochs (set valid=valid.txt or train.txt in obj.data file) and run: darknet.exe detector train data/obj.data yolo-obj.cfg darknet53.conv.74 -map

  1. After training is complete - get result yolo-obj_final.weights from path build\darknet\x64\backup\

  • After each 100 iterations you can stop and later start training from this point. For example, after 2000 iterations you can stop training, and later just copy yolo-obj_2000.weights from build\darknet\x64\backup\ to build\darknet\x64\ and start training using: darknet.exe detector train data/obj.data yolo-obj.cfg yolo-obj_2000.weights

    (in the original repository https://github.com/pjreddie/darknet the weights-file is saved only once every 10 000 iterations if(iterations > 1000))

  • Also you can get result earlier than all 45000 iterations.

Note: If during training you see nan values for avg (loss) field - then training goes wrong, but if nan is in some other lines - then training goes well.

Note: If you changed width= or height= in your cfg-file, then new width and height must be divisible by 32.

Note: After training use such command for detection: darknet.exe detector test data/obj.data yolo-obj.cfg yolo-obj_8000.weights

Note: if error Out of memory occurs then in .cfg-file you should increase subdivisions=16, 32 or 64: link

How to train tiny-yolo (to detect your custom objects):

Do all the same steps as for the full yolo model as described above. With the exception of:

  • Download default weights file for yolov3-tiny: https://pjreddie.com/media/files/yolov3-tiny.weights
  • Get pre-trained weights yolov3-tiny.conv.15 using command: darknet.exe partial cfg/yolov3-tiny.cfg yolov3-tiny.weights yolov3-tiny.conv.15 15
  • Make your custom model yolov3-tiny-obj.cfg based on cfg/yolov3-tiny_obj.cfg instead of yolov3.cfg
  • Start training: darknet.exe detector train data/obj.data yolov3-tiny-obj.cfg yolov3-tiny.conv.15

For training Yolo based on other models (DenseNet201-Yolo or ResNet50-Yolo), you can download and get pre-trained weights as showed in this file: https://github.com/AlexeyAB/darknet/blob/master/build/darknet/x64/partial.cmd If you made you custom model that isn't based on other models, then you can train it without pre-trained weights, then will be used random initial weights.

When should I stop training:

Usually sufficient 2000 iterations for each class(object), but not less than 4000 iterations in total. But for a more precise definition when you should stop training, use the following manual:

  1. During training, you will see varying indicators of error, and you should stop when no longer decreases 0.XXXXXXX avg:

Region Avg IOU: 0.798363, Class: 0.893232, Obj: 0.700808, No Obj: 0.004567, Avg Recall: 1.000000, count: 8 Region Avg IOU: 0.800677, Class: 0.892181, Obj: 0.701590, No Obj: 0.004574, Avg Recall: 1.000000, count: 8

9002: 0.211667, 0.060730 avg, 0.001000 rate, 3.868000 seconds, 576128 images Loaded: 0.000000 seconds

  • 9002 - iteration number (number of batch)
  • 0.060730 avg - average loss (error) - the lower, the better

When you see that average loss 0.xxxxxx avg no longer decreases at many iterations then you should stop training.

  1. Once training is stopped, you should take some of last .weights-files from darknet\build\darknet\x64\backup and choose the best of them:

For example, you stopped training after 9000 iterations, but the best result can give one of previous weights (7000, 8000, 9000). It can happen due to overfitting. Overfitting - is case when you can detect objects on images from training-dataset, but can't detect objects on any others images. You should get weights from Early Stopping Point:

Overfitting

To get weights from Early Stopping Point:

2.1. At first, in your file obj.data you must specify the path to the validation dataset valid = valid.txt (format of valid.txt as in train.txt), and if you haven't validation images, just copy data\train.txt to data\valid.txt.

2.2 If training is stopped after 9000 iterations, to validate some of previous weights use this commands:

(If you use another GitHub repository, then use darknet.exe detector recall... instead of darknet.exe detector map...)

  • darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_7000.weights
  • darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_8000.weights
  • darknet.exe detector map data/obj.data yolo-obj.cfg backup\yolo-obj_9000.weights

And comapre last output lines for each weights (7000, 8000, 9000):

Choose weights-file with the highest mAP (mean average precision) or IoU (intersect over union)

For example, bigger mAP gives weights yolo-obj_8000.weights - then use this weights for detection.

Or just train with -map flag:

darknet.exe detector train data/obj.data yolo-obj.cfg darknet53.conv.74 -map

So you will see mAP-chart (red-line) in the Loss-chart Window. mAP will be calculated for each 4 Epochs using valid=valid.txt file that is specified in obj.data file (1 Epoch = images_in_train_txt / batch iterations)

loss_chart_map_chart

Example of custom object detection: darknet.exe detector test data/obj.data yolo-obj.cfg yolo-obj_8000.weights

  • IoU (intersect over union) - average instersect over union of objects and detections for a certain threshold = 0.24

  • mAP (mean average precision) - mean value of average precisions for each class, where average precision is average value of 11 points on PR-curve for each possible threshold (each probability of detection) for the same class (Precision-Recall in terms of PascalVOC, where Precision=TP/(TP+FP) and Recall=TP/(TP+FN) ), page-11: http://homepages.inf.ed.ac.uk/ckiw/postscript/ijcv_voc09.pdf

mAP is default metric of precision in the PascalVOC competition, this is the same as AP50 metric in the MS COCO competition. In terms of Wiki, indicators Precision and Recall have a slightly different meaning than in the PascalVOC competition, but IoU always has the same meaning.

precision_recall_iou

How to calculate mAP on PascalVOC 2007:

  1. To calculate mAP (mean average precision) on PascalVOC-2007-test:

(The article specifies the value of mAP = 76.8% for YOLOv2 416×416, page-4 table-3: https://arxiv.org/pdf/1612.08242v1.pdf. We get values lower - perhaps due to the fact that the model was trained on a slightly different source code than the code on which the detection is was done)

  • if you want to get mAP for tiny-yolo-voc.cfg model, then un-comment line for tiny-yolo-voc.cfg and comment line for yolo-voc.cfg in the .cmd-file
  • if you have Python 2.x instead of Python 3.x, and if you use Darknet+Python-way to get mAP, then in your cmd-file use reval_voc.py and voc_eval.py instead of reval_voc_py3.py and voc_eval_py3.py from this directory: https://github.com/AlexeyAB/darknet/tree/master/scripts

Custom object detection:

Example of custom object detection: darknet.exe detector test data/obj.data yolo-obj.cfg yolo-obj_8000.weights

Yolo_v2_training Yolo_v2_training

How to improve object detection:

  1. Before training:

  • set flag random=1 in your .cfg-file - it will increase precision by training Yolo for different resolutions: link

  • increase network resolution in your .cfg-file (height=608, width=608 or any value multiple of 32) - it will increase precision

  • recalculate anchors for your dataset for width and height from cfg-file: darknet.exe detector calc_anchors data/obj.data -num_of_clusters 9 -width 416 -height 416 then set the same 9 anchors in each of 3 [yolo]-layers in your cfg-file

  • check that each object are mandatory labeled in your dataset - no one object in your data set should not be without label. In the most training issues - there are wrong labels in your dataset (got labels by using some conversion script, marked with a third-party tool, ...). Always check your dataset by using: https://github.com/AlexeyAB/Yolo_mark

  • desirable that your training dataset include images with objects at diffrent: scales, rotations, lightings, from different sides, on different backgrounds - you should preferably have 2000 different images for each class or more, and you should train 2000*classes iterations or more

  • desirable that your training dataset include images with non-labeled objects that you do not want to detect - negative samples without bounded box (empty .txt files) - use as many images of negative samples as there are images with objects

  • for training with a large number of objects in each image, add the parameter max=200 or higher value in the last [yolo]-layer or [region]-layer in your cfg-file (the global maximum number of objects that can be detected by YoloV3 is 0,0615234375*(width*height) where are width and height are parameters from [net] section in cfg-file)

  • for training for small objects - set layers = -1, 11 instead of 6390a5a2ab/cfg/yolov3.cfg (L720) and set stride=4 instead of 6390a5a2ab/cfg/yolov3.cfg (L717)

  • If you train the model to distinguish Left and Right objects as separate classes (left/right hand, left/right-turn on road signs, ...) then for disabling flip data augmentation - add flip=0 here: 3d2d0a7c98/cfg/yolov3.cfg (L17)

  • General rule - your training dataset should include such a set of relative sizes of objects that you want to detect:

    • train_network_width * train_obj_width / train_image_width ~= detection_network_width * detection_obj_width / detection_image_width
    • train_network_height * train_obj_height / train_image_height ~= detection_network_height * detection_obj_height / detection_image_height

    I.e. for each object from Test dataset there must be at least 1 object in the Training dataset with about the same relative size:

    object width in percent from Training dataset ~= object width in percent from Test dataset

    That is, if only objects that occupied 80-90% of the image were present in the training set, then the trained network will not be able to detect objects that occupy 1-10% of the image.

  • to speedup training (with decreasing detection accuracy) do Fine-Tuning instead of Transfer-Learning, set param stopbackward=1 here: 6d44529cf9/cfg/yolov3.cfg (L548) then do this command: ./darknet partial cfg/yolov3.cfg yolov3.weights yolov3.conv.81 81 will be created file yolov3.conv.81, then train by using weights file yolov3.conv.81 instead of darknet53.conv.74

  1. After training - for detection:

  • Increase network-resolution by set in your .cfg-file (height=608 and width=608) or (height=832 and width=832) or (any value multiple of 32) - this increases the precision and makes it possible to detect small objects: link

    • it is not necessary to train the network again, just use .weights-file already trained for 416x416 resolution
    • but to get even greater accuracy you should train with higher resolution 608x608 or 832x832, note: if error Out of memory occurs then in .cfg-file you should increase subdivisions=16, 32 or 64: link

How to mark bounded boxes of objects and create annotation files:

Here you can find repository with GUI-software for marking bounded boxes of objects and generating annotation files for Yolo v2 & v3: https://github.com/AlexeyAB/Yolo_mark

With example of: train.txt, obj.names, obj.data, yolo-obj.cfg, air1-6.txt, bird1-4.txt for 2 classes of objects (air, bird) and train_obj.cmd with example how to train this image-set with Yolo v2 & v3

Using Yolo9000

Simultaneous detection and classification of 9000 objects: darknet.exe detector test cfg/combine9k.data cfg/yolo9000.cfg yolo9000.weights data/dog.jpg

How to use Yolo as DLL

  1. To compile Yolo as C++ DLL-file yolo_cpp_dll.dll - open in MSVS2015 file build\darknet\yolo_cpp_dll.sln, set x64 and Release, and do the: Build -> Build yolo_cpp_dll

    • You should have installed CUDA 9.1
    • To use cuDNN do: (right click on project) -> properties -> C/C++ -> Preprocessor -> Preprocessor Definitions, and add at the beginning of line: CUDNN;

  2. To use Yolo as DLL-file in your C++ console application - open in MSVS2015 file build\darknet\yolo_console_dll.sln, set x64 and Release, and do the: Build -> Build yolo_console_dll

    • you can run your console application from Windows Explorer build\darknet\x64\yolo_console_dll.exe use this command: yolo_console_dll.exe data/coco.names yolov3.cfg yolov3.weights test.mp4

    • or you can run from MSVS2015 (before this - you should copy 2 files yolo-voc.cfg and yolo-voc.weights to the directory build\darknet\ )

    • after launching your console application and entering the image file name - you will see info for each object: <obj_id> <left_x> <top_y> <width> <height> <probability>

    • to use simple OpenCV-GUI you should uncomment line //#define OPENCV in yolo_console_dll.cpp-file: link

    • you can see source code of simple example for detection on the video file: link

yolo_cpp_dll.dll-API: link

class Detector {
public:
	Detector(std::string cfg_filename, std::string weight_filename, int gpu_id = 0);
	~Detector();

	std::vector<bbox_t> detect(std::string image_filename, float thresh = 0.2, bool use_mean = false);
	std::vector<bbox_t> detect(image_t img, float thresh = 0.2, bool use_mean = false);
	static image_t load_image(std::string image_filename);
	static void free_image(image_t m);

#ifdef OPENCV
	std::vector<bbox_t> detect(cv::Mat mat, float thresh = 0.2, bool use_mean = false);
#endif
};