yolov5目标检测模型的知识蒸馏（基于响应的蒸馏）

Last update: Dec 04, 2022

Overview

代码地址：

https://github.com/Sharpiless/yolov5-knowledge-distillation

教师模型：

python train.py --weights weights/yolov5m.pt \
        --cfg models/yolov5m.yaml --data data/voc.yaml --epochs 50 \
        --batch-size 8 --device 0 --hyp data/hyp.scratch.yaml

蒸馏训练：

python train.py --weights weights/yolov5s.pt \
        --cfg models/yolov5s.yaml --data data/voc.yaml --epochs 50 \
        --batch-size 8 --device 0 --hyp data/hyp.scratch.yaml \
        --t_weights yolov5m.pt --distill

训练参数:

--weights：预训练模型

--t_weights：教师模型权重

--distill：使用知识蒸馏进行训练

--dist_loss：l2或者kl

--temperature：使用知识蒸馏时的温度

使用《Object detection at 200 Frames Per Second》中的损失

这篇文章分别对这几个损失函数做出改进，具体思路为只有当teacher network的objectness value高时，才学习bounding box坐标和class probabilities。

实验结果：

这里假设VOC2012中新增加的数据为无标签数据（2k张）。

教师模型	训练方法	蒸馏损失	P	R	mAP50
无	正常训练	不使用	0.7756	0.7115	0.7609
Yolov5l	output based	l2	0.7585	0.7198	0.7644
Yolov5l	output based	KL	0.7417	0.7207	0.7536
Yolov5m	output based	l2	0.7682	0.7436	0.7976
Yolov5m	output based	KL	0.7731	0.7313	0.7931

参数和细节正在完善，支持KL散度、L2 logits损失和Sigmoid蒸馏损失等

1. 正常训练：

2. L2蒸馏损失：

我的公众号：

关于作者

B站：https://space.bilibili.com/470550823

CSDN：https://blog.csdn.net/weixin_44936889

AI Studio：https://aistudio.baidu.com/aistudio/personalcenter/thirdview/67156

Github：https://github.com/Sharpiless

yolov5目标检测模型的知识蒸馏（基于响应的蒸馏）

Related tags

Overview

代码地址：

教师模型：

蒸馏训练：

训练参数:

实验结果：

1. 正常训练：

2. L2蒸馏损失：

我的公众号：

关于作者

Owner

X-modaler is a versatile and high-performance codebase for cross-modal analytics.

Utility code for use with PyXLL

Unsupervised phone and word segmentation using dynamic programming on self-supervised VQ features.

The PASS dataset: pretrained models and how to get the data - PASS: Pictures without humAns for Self-Supervised Pretraining

Implementation of Deformable Attention in Pytorch from the paper "Vision Transformer with Deformable Attention"

YOLTv4 builds upon YOLT and SIMRDWN, and updates these frameworks to use the most performant version of YOLO, YOLOv4

Breaking the Curse of Space Explosion: Towards Efficient NAS with Curriculum Search

Implementation of Convolutional LSTM in PyTorch.

This repository is all about spending some time the with the original problem posed by Minsky and Papert

LieTransformer: Equivariant Self-Attention for Lie Groups

[ICCV'21] Neural Radiance Flow for 4D View Synthesis and Video Processing

Implementation of Online Label Smoothing in PyTorch

Robustness via Cross-Domain Ensembles

I will implement Fastai in each projects present in this repository.

Official code of our work, AVATAR: A Parallel Corpus for Java-Python Program Translation.

Repository of the paper Compressing Sensor Data for Remote Assistance of Autonomous Vehicles using Deep Generative Models at ML4AD @ NeurIPS 2021.

This game was designed to encourage young people not to gamble on lotteries, as the probablity of correctly guessing the number is infinitesimal!

NeRF Meta-Learning with PyTorch

Face recognition. Redefined.

Implementation of H-UCRL Algorithm