pytorch, hand(object) detect ,yolo v5，手检测

Last update: Dec 20, 2022

Related tags

Deep Learning yolo-v5

Overview

YOLO V5

物体检测，包括手部检测。

项目介绍

手部检测

手部检测示例如下：

视频示例：

项目配置

作者开发环境：
Python 3.7
PyTorch >= 1.5.1

数据集

手部检测数据集

该项目数据集采用 TV-Hand 和 COCO-Hand (COCO-Hand-Big 部分) 进行制作。
TV-Hand 和 COCO-Hand数据集官网地址 http://vision.cs.stonybrook.edu/~supreeth/

感谢数据集贡献者。    
Paper：  
Contextual Attention for Hand Detection in the Wild. S. Narasimhaswamy, Z. Wei, Y. Wang, J. Zhang, and M. Hoai, IEEE International Conference on Computer Vision, ICCV 2019.

该项目制作的训练集的数据集下载地址(百度网盘 Password: 25y3 )

所有数据集的数据格式

size是全图分辨率， (x，y) 是目标物体中心对于全图的归一化坐标，w,h是目标物体边界框对于全图的归一化宽、高。

dw = 1./(size[0])  
dh = 1./(size[1])  
x = (box[0] + box[1])/2.0 - 1  
y = (box[2] + box[3])/2.0 - 1  
w = box[1] - box[0]  
h = box[3] - box[2]  
x = x*dw  
w = w*dw  
y = y*dh  
h = h*dh

为了更好了解标注数据格式，可以通过运行 show_yolo_anno.py 脚本进行制作数据集的格式。注意配置脚本里的path和path_voc_names，path为标注数据集的相关文件路径，path_voc_names为数据集配置文件。

制作自己的训练数据集

如下所示,每一行代表一个物体实例，第一列是标签，后面是归一化的中心坐标(x,y),和归一化的宽(w)和高(h)，且每一列信息空格间隔。归一化公式如上，同时可以通过show_yolo_anno.py进行参数适配后，可视化验证其正确性。

label     x                  y                   w                  h
0 0.6200393316313977 0.5939000244140625 0.17241466452130497 0.14608001708984375
0 0.38552491996544863 0.5855700073242187 0.14937006832733554 0.1258599853515625
0 0.32889763138738515 0.701989990234375 0.031338589085055775 0.0671400146484375
0 0.760577424617577 0.69422998046875 0.028556443261975064 0.0548599853515625
0 0.5107086662232406 0.6921500244140625 0.018792660530470802 0.04682000732421875
0 0.9295538153861138 0.67602001953125 0.03884511231750328 0.01844000244140625

预训练模型

从零开始预训练模型

预训练模型下载地址(百度网盘 Password: ad4l )

手部检测预训练模型

包括yolo_v5预训练模型图像输入尺寸640。
预训练模型下载地址(百度网盘 Password: x7d4 )

项目使用方法

数据集可视化

根目录下运行命令： show_yolo_anno.py (注意脚本内相关参数配置 )

模型训练

根目录下运行命令： python train.py (注意脚本内相关参数配置 )

模型推理

根目录下运行命令： python video.py (注意脚本内相关参数配置 )

pytorch, hand(object) detect ,yolo v5，手检测

Related tags

Overview

YOLO V5

项目介绍

手部检测

项目配置

数据集

手部检测数据集

所有数据集的数据格式

制作自己的训练数据集

预训练模型

从零开始预训练模型

手部检测预训练模型

项目使用方法

数据集可视化

模型训练

模型推理

Owner

Eric.Lee

Python scripts for performing stereo depth estimation using the MobileStereoNet model in Tensorflow Lite.

Code for "Offline Meta-Reinforcement Learning with Advantage Weighting" [ICML 2021]

PyTorch Implementation of Region Similarity Representation Learning (ReSim)

Code repository for our paper regarding the L3D dataset.

A PaddlePaddle implementation of STGCN with a few modifications in the model architecture in order to forecast traffic jam.

This is an official implementation of our CVPR 2021 paper "Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regression" (https://arxiv.org/abs/2104.02300)

[NeurIPS'21] "AugMax: Adversarial Composition of Random Augmentations for Robust Training" by Haotao Wang, Chaowei Xiao, Jean Kossaifi, Zhiding Yu, Animashree Anandkumar, and Zhangyang Wang.

[CVPR 2021] MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition

The story of Chicken for Club Bing

Official code for UnICORNN (ICML 2021)

Cycle Consistent Adversarial Domain Adaptation (CyCADA)

Subnet Replacement Attack: Towards Practical Deployment-Stage Backdoor Attack on Deep Neural Networks

CBKH: The Cornell Biomedical Knowledge Hub

Simultaneous NMT/MMT framework in PyTorch

MADE (Masked Autoencoder Density Estimation) implementation in PyTorch

Code of PVTv2 is released! PVTv2 largely improves PVTv1 and works better than Swin Transformer with ImageNet-1K pre-training.

A transformer which can randomly augment VOC format dataset (both image and bbox) online.

Stratified Transformer for 3D Point Cloud Segmentation (CVPR 2022)

This is RFA-Toolbox, a simple and easy-to-use library that allows you to optimize your neural network architectures using receptive field analysis (RFA) and create graph visualizations of your architecture.

Save-restricted-v-3 - Save restricted content Bot For telegram