Relation Knowledge Distillation by Auxiliary Learning for Object Detection

关系(数据库) 计算机科学 蒸馏 目标检测 人工智能 任务(项目管理) 管道(软件) 对象(语法) 推论 模式识别(心理学) 相关性(法律) 机器学习 数据挖掘 工程类 化学 有机化学 系统工程 法学 政治学 程序设计语言
作者
Hao Wang,Tong Jia,Qilong Wang,Wangmeng Zuo
出处
期刊:IEEE transactions on image processing [Institute of Electrical and Electronics Engineers]
卷期号:33: 4796-4810 被引量:15
标识
DOI:10.1109/tip.2024.3445740
摘要

Balancing the trade-off between accuracy and speed for obtaining higher performance without sacrificing the inference time is a challenging topic for object detection task. Knowledge distillation, which serves as a kind of model compression techniques, provides a potential and feasible way to handle above efficiency and effectiveness issue through transferring the dark knowledge from the sophisticated teacher detector to the simple student one. Despite demonstrating promising solutions to make harmonies between accuracy and speed, current knowledge distillation for object detection methods still suffer from two limitations. Firstly, most of the methods are inherited or refereed from the frameworks in image classification task, and deploy an implicit manner by imitating or constraining the features from the intermediate layers or the output predictions between the teacher and student models. While little consideration has been raised to the intrinsic relevance of the classification and localization predictions in object detection task. Besides, these methods fail to investigate the relationship between detection and distillation tasks in knowledge distillation pipeline, and they train the whole network by simply integrating losses from these two different tasks through hand-crafted designation parameters. For addressing the aforementioned issues, we propose a novel Relation Knowledge Distillation by Auxiliary Learning for Object Detection (ReAL) method in this paper. Specifically, we first design a prediction relation distillation module which makes the student model directly mimic the output predictions from the teacher one, and conduct self and mutual relation distillation losses to excavate the relation information between teacher and student models. Moreover, for better devolving into the relationship between different tasks in distillation pipeline, we introduce the auxiliary learning into knowledge distillation for object detection and develop a dynamic weight adaptation strategy. Through regarding detection task as primary task and treating distillation task as auxiliary task in auxiliary learning framework, we dynamically adjust and regularize the corresponding weights of the losses for these tasks during the training process. Experiments on MS COCO dataset are conducted using various detector combinations of teacher and student models and the results show that our proposed ReAL can achieve obvious improvement on different distillation model configurations, while performing favorably against state-of-the-arts.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
舒适沛蓝发布了新的文献求助10
刚刚
刚刚
刚刚
漠池完成签到,获得积分10
刚刚
领导范儿的应助被嘎嘣脆采纳,获得10
1秒前
1秒前
Jun发布了新的文献求助10
2秒前
tt发布了新的文献求助10
2秒前
愤怒也哈哈完成签到,获得积分10
3秒前
3秒前
4秒前
5秒前
5秒前
Z20251040501发布了新的文献求助10
5秒前
安静冷菱关注了科研通微信公众号
6秒前
6秒前
小二郎的应助被akanenn999采纳,获得10
6秒前
斯文败类的应助被辛勤的谷冬采纳,获得10
6秒前
Sennir关注了科研通微信公众号
7秒前
一哈哈发布了新的文献求助10
7秒前
8秒前
8秒前
TFauthor发布了新的文献求助10
9秒前
9秒前
科研通AI6.2的应助被zxd采纳,获得10
9秒前
张祖伦发布了新的文献求助10
9秒前
10秒前
11秒前
crrrr完成签到,获得积分20
11秒前
Tina完成签到,获得积分10
11秒前
xhzdd发布了新的文献求助10
11秒前
ding的应助被yw采纳,获得10
12秒前
12秒前
晓宇发布了新的文献求助10
12秒前
12秒前
lili发布了新的文献求助10
13秒前
科研通AI6.4的应助被ddd采纳,获得10
13秒前
元昭诩发布了新的文献求助10
13秒前
果子爱学习完成签到 ,获得积分10
13秒前
妇产科医生完成签到,获得积分10
13秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
八旗制度の研究 600
高分子基功能界面水蒸发材料与应用研究 500
外场增强的水汽发电研究 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8187918
求助须知:如何正确求助?哪些是违规求助? 9575850
关于积分的说明 21843829
捐赠科研通 7785760
什么是DOI,文献DOI怎么找? 3435128
关于科研通互助平台的介绍 2559501
邀请新用户注册赠送积分活动 2482027