All Projects → DataXujing → DIoU_YOLO_V3

DataXujing / DIoU_YOLO_V3

Licence: GPL-3.0 license
📈📈📈【口罩佩戴检测数据训练 | 开源口罩检测数据集和预训练模型】Train D/CIoU_YOLO_V3 by darknet for object detection

Programming Languages

Jupyter Notebook
11667 projects
c
50402 projects - #5 most used programming language
python
139335 projects - #7 most used programming language
matlab
3953 projects
Cuda
1817 projects
C++
36643 projects - #6 most used programming language

Projects that are alternatives of or similar to DIoU YOLO V3

Mobilenet Yolo
MobileNetV2-YoloV3-Nano: 0.5BFlops 3MB HUAWEI P40: 6ms/img, YoloFace-500k:0.1Bflops 420KB🔥🔥🔥
Stars: ✭ 1,566 (+2854.72%)
Mutual labels:  cv, darknet, yolov3
Bmw Yolov4 Training Automation
This repository allows you to get started with training a state-of-the-art Deep Learning model with little to no configuration needed! You provide your labeled dataset or label your dataset using our BMW-LabelTool-Lite and you can start the training right away and monitor it in many different ways like TensorBoard or a custom REST API and GUI. NoCode training with YOLOv4 and YOLOV3 has never been so easy.
Stars: ✭ 533 (+905.66%)
Mutual labels:  darknet, yolov3
Yolo3 4 Py
A Python wrapper on Darknet. Compatible with YOLO V3.
Stars: ✭ 504 (+850.94%)
Mutual labels:  darknet, yolov3
Pytorch Onnx Tensorrt
A set of tool which would make your life easier with Tensorrt and Onnxruntime. This Repo is designed for YoloV3
Stars: ✭ 66 (+24.53%)
Mutual labels:  darknet, yolov3
Tensorflow 2.x Yolov3
YOLOv3 implementation in TensorFlow 2.3.1
Stars: ✭ 300 (+466.04%)
Mutual labels:  darknet, yolov3
Msnhnet
🔥 (yolov3 yolov4 yolov5 unet ...)A mini pytorch inference framework which inspired from darknet.
Stars: ✭ 357 (+573.58%)
Mutual labels:  darknet, yolov3
Yolo annotation tool
Annotation tool for YOLO in opencv
Stars: ✭ 17 (-67.92%)
Mutual labels:  darknet, yolov3
object-tracking
Multiple Object Tracking System in Keras + (Detection Network - YOLO)
Stars: ✭ 89 (+67.92%)
Mutual labels:  darknet, yolov3
Pytorch Imagenet Cifar Coco Voc Training
Training examples and results for ImageNet(ILSVRC2012)/CIFAR100/COCO2017/VOC2007+VOC2012 datasets.Image Classification/Object Detection.Include ResNet/EfficientNet/VovNet/DarkNet/RegNet/RetinaNet/FCOS/CenterNet/YOLOv3.
Stars: ✭ 130 (+145.28%)
Mutual labels:  darknet, yolov3
Darknet face with landmark
加入关键点的darknet训练框架,轻量级的人脸检测,支持ncnn推理
Stars: ✭ 192 (+262.26%)
Mutual labels:  darknet, yolov3
Pine
🌲 Aimbot powered by real-time object detection with neural networks, GPU accelerated with Nvidia. Optimized for use with CS:GO.
Stars: ✭ 202 (+281.13%)
Mutual labels:  darknet, yolov3
car-crash-accident
Car Crash Accident Project
Stars: ✭ 14 (-73.58%)
Mutual labels:  darknet, yolov3
drone-net
https://towardsdatascience.com/tutorial-build-an-object-detection-system-using-yolo-9a930513643a
Stars: ✭ 126 (+137.74%)
Mutual labels:  darknet, yolov3
Yolo Tensorrt
Support Yolov5s,m,l,x .darknet -> tensorrt. Yolov4 Yolov3 use raw darknet *.weights and *.cfg fils. If the wrapper is useful to you,please Star it.
Stars: ✭ 495 (+833.96%)
Mutual labels:  darknet, yolov3
live-cctv
To detect any reasonable change in a live cctv to avoid large storage of data. Once, we notice a change, our goal would be track that object or person causing it. We would be using Computer vision concepts. Our major focus will be on Deep Learning and will try to add as many features in the process.
Stars: ✭ 23 (-56.6%)
Mutual labels:  darknet, yolov3
Mobilenet Yolo
A caffe implementation of MobileNet-YOLO detection network
Stars: ✭ 825 (+1456.6%)
Mutual labels:  darknet, yolov3
Deep-Learning-with-GoogleColab
Deep Learning Applications (Darknet - YOLOv3, YOLOv4 | DeOldify - Image Colorization, Video Colorization | Face-Recognition) with Google Colaboratory - on the free Tesla K80/Tesla T4/Tesla P100 GPU - using Keras, Tensorflow and PyTorch.
Stars: ✭ 63 (+18.87%)
Mutual labels:  darknet, yolov3
odam
ODAM - Object detection and Monitoring
Stars: ✭ 16 (-69.81%)
Mutual labels:  darknet, yolov3
Darknet2caffe
Convert darknet weights to caffemodel
Stars: ✭ 127 (+139.62%)
Mutual labels:  darknet, yolov3
Yolo person detect
person detect based on yolov3 with several Python scripts
Stars: ✭ 212 (+300%)
Mutual labels:  darknet, yolov3

DIoU-Darknet 口罩佩戴检测

提供整理的开源口罩佩戴检测数据集的下载!!!

检测效果 好于AIZOO https://aizoo.com/face-mask-detection.html

检测效果 好于百度开源口罩检测 https://www.paddlepaddle.org.cn/hub/scene/maskdetect

YOLOv3 with DIoU and CIoU losses implemented in Darknet

[arxiv] [pdf]

Xu Jing

DEMO

相同图片基于AIZOO开源的检测效果:

配置和训练过程以darknet版YOLO v3和Gaussian YOLO v3的配置训练过程相似,这里以Ubuntu为例,关于Windows下的配置和生成过程,具体可以联系DataXujing

1.修改Makefile并make

GPU=1 #如果使用GPU设置为1,CPU设置为0
CUDNN=1  #如果使用CUDNN设置为1,否则为0
OPENCV=1 #如果调用摄像头,还需要设置OPENCV为1,否则为0
OPENMP=1  #如果使用OPENMP设置为1,否则为0
DEBUG=0  #如果使用DEBUG设置为1,否则为0
make

2.数据来源

数据整理成VOC格式的数据,最终数据为8565张图像,由于存储空间问题,如果开发者需要改数据,请通过个人GitHub主页公布的邮箱索要。

将数据存放在data/myData

3.配置文件

  • (1).Losses

目前支持的损失:[iou|giou|diou|ciou|mse]

iou_loss=ciou
  • (2).Normalizers

Location和Claffication损失之间的Normalize,cls_normalizeriou_normalizer 默认都是1.0,但在DIoU探索中发现,下面的配置比较好:

iou_loss=diou
cls_normalizer=1
iou_normalizer=1.0
iou_loss=ciou
cls_normalizer=1
iou_normalizer=0.5
  • (3).DIoU-NMS

nms_kind参数下设置,目前支持[greedynms|diounms]

nms_kind=greedynms
nms_kind=diounms

对于YOLOv3来说, DIoU-NMS中引入beta1参数: DIoU = IoU - R_DIoU ^ {beta1},是的NMS表现不错。DIoU发现,对于YOLOv3

beta1=0.6

DIoU-NMS是比较好的(bet1=1.0也优于greedy-NMS)。但对于SSD和Faster R-CNN而言,beta1=1.0就很不错了。

  • (4).learning rate

    • learning_rate的设置原则 NEW_RATE = ORIGINAL_RATE * 1/NUMBER_OF_GPUS

    • burn_in 的设置原则 NEW_BURN_IN = ORIGINAL_BURN_IN * NUMBER_OF_GPUS

因此对于1个GPU的情形:

    learning_rate=0.001
    burn_in=1000

对于2个GPU的情形:

    learning_rate=0.0005
    burn_in=2000

对于4个GPU的情形:

    learning_rate=0.00025
    burn_in=4000
  • (5).其他配置

其他配置和YOLOv3相似!

4.训练

./darknet detector train cfg/myData-diou.data cfg/myData-diou.cfg darknet53.conv.74 #-gpus 0,1

5.Evaluation

./darknet detector test cfg/myData-ciou.data cfg/myData-ciou.cfg backup/ciou/myData-ciou_14000.weights data/test1.jpg

Reference

@inproceedings{zheng2020distance,
  author    = {Zhaohui Zheng, Ping Wang, Wei Liu, Jinze Li, Rongguang Ye, Dongwei Ren},
  title     = {Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression},
  booktitle = {The AAAI Conference on Artificial Intelligence (AAAI)},
   year      = {2020},
}
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].