GitHub - zengwb-lx/Yolov5-Deepsort-Fastreid

详细请看博客：https://blog.csdn.net/zengwubbb/article/details/113422048 reid模型为fast-reid框架训练，resnet101蒸馏出来的resnet34，因为模型保存了FC层和优化器等参数所以很大，去除这些只保留resnet34的话模型30多MB，整个流程在2070GPU下能达到实时效果 reid模型：链接: https://pan.baidu.com/s/1bMG3qy7npecCh6AzNO-Zyw 提取码: hy1m

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
configs		configs
deep_sort		deep_sort
fast_reid		fast_reid
kd-r34-r101_ibn		kd-r34-r101_ibn
models		models
pedestrians-cluster		pedestrians-cluster
utils		utils
weights		weights
.dockerignore		.dockerignore
.gitattributes		.gitattributes
.gitignore		.gitignore
Dockerfile		Dockerfile
LICENSE		LICENSE
README.md		README.md
detect.py		detect.py
hubconf.py		hubconf.py
person_count.py		person_count.py
person_detect_yolov5.py		person_detect_yolov5.py
person_search_reid.py		person_search_reid.py
requirements.txt		requirements.txt
test.py		test.py
train.py		train.py
tutorial.ipynb		tutorial.ipynb

Provide feedback