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

Official implementation of Meta-StyleSpeech and StyleSpeech

Generative Handwriting using LSTM Mixture Density Network with TensorFlow

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

Expressive Power of Invariant and Equivaraint Graph Neural Networks (ICLR 2021)

Official codebase for Legged Robots that Keep on Learning: Fine-Tuning Locomotion Policies in the Real World

Pytorch implementation of "Training a 85.4% Top-1 Accuracy Vision Transformer with 56M Parameters on ImageNet"

Implementation for HFGI: High-Fidelity GAN Inversion for Image Attribute Editing

HAT: Hierarchical Aggregation Transformers for Person Re-identification

codes for IKM (arXiv2021, Submitted to IEEE Trans)

Learning Chinese Character style with conditional GAN

This repository contains the database and code used in the paper Embedding Arithmetic for Text-driven Image Transformation

Competitive Programming Club, Clinify's Official repository for CP problems hosting by club members.

Differentiable rasterization applied to 3D model simplification tasks

Deep Convolutional Generative Adversarial Networks

Generative Models for Graph-Based Protein Design

This is the repository of the NeurIPS 2021 paper "Curriculum Disentangled Recommendation withNoisy Multi-feedback"

Attention Probe: Vision Transformer Distillation in the Wild

Spatially-Adaptive Pixelwise Networks for Fast Image Translation, CVPR 2021

FLAVR is a fast, flow-free frame interpolation method capable of single shot multi-frame prediction

HackBMU-5.0-Team-Ctrl-Alt-Elite - HackBMU 5.0 Team Ctrl Alt Elite