CASIA OpenIR  > 模式识别国家重点实验室  > 图像与视频分析
Large Batch Optimization for Object Detection: Training COCO in 12 minutes
Wang, Tong1,2; Zhu, Yousong1,3; Zhao, Chaoyang1; Zeng, Wei4,5; Wang, Yaowei5; Wang, Jinqiao1,2,6; Tang, Ming1
2020-08
Conference NameEuropean Conference on Compter Vision (ECCV)
Pages481-496
Conference Date2020-8-24
Conference PlaceOnline
Abstract

Most of existing object detectors usually adopt a small training batch size (e.g. 16), which severely hinders the whole community from exploring large-scale datasets due to the extremely long training procedure. In this paper, we propose a versatile large batch optimization framework for object detection, named LargeDet, which successfully scales the batch size to larger than 1K for the first time. Specifically, we present a novel Periodical Moments Decay LAMB (PMD-LAMB) algorithm to effectively reduce the negative effects of the lagging historical gradients. Additionally, the Synchronized Batch Normalization (SyncBN) is utilized to help fast convergence. With LargeDet, we can not only prominently shorten the training period, but also significantly improve the detection accuracy of sparsely annotated large-scale datasets. For instance, we can finish the training of ResNet50 FPN detector on COCO within 12 minutes. Moreover, we achieve 12.2% mAP@0.5 absolute improvement for ResNet50 FPN on Open Images by training with batch size 640.

KeywordObject detection Large batch optimization Periodical moments decay
Subject Area模式识别
MOST Discipline Catalogue工学::计算机科学与技术(可授工学、理学学位)
Indexed ByEI
Funding ProjectNational Natural Science Foundation of China (NSFC)[61633002] ; National Nature Science Foundation of China[61876086] ; National Natural Science Foundation of China[61702510] ; National Natural Science Foundation of China[61806200] ; National Natural Science Foundation of China[61772527] ; National Natural Science Foundation of China[61772527] ; National Natural Science Foundation of China[61806200] ; National Natural Science Foundation of China[61702510] ; National Nature Science Foundation of China[61876086] ; National Natural Science Foundation of China (NSFC)[61633002]
Language英语
Document Type会议论文
Identifierhttp://ir.ia.ac.cn/handle/173211/47416
Collection模式识别国家重点实验室_图像与视频分析
Affiliation1.National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing, China
2.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China
3.ObjectEye Inc., Beijing, China
4.Peking University, Beijing, China
5.Peng Cheng Laboratory, Shenzhen, China
6.NEXWISE Co., Ltd, Guangzhou, China
First Author AffilicationChinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
Recommended Citation
GB/T 7714
Wang, Tong,Zhu, Yousong,Zhao, Chaoyang,et al. Large Batch Optimization for Object Detection: Training COCO in 12 minutes[C],2020:481-496.
Files in This Item: Download All
File Name/Size DocType Version Access License
Wang2020_Chapter_Lar(3706KB)会议论文 开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Wang, Tong]'s Articles
[Zhu, Yousong]'s Articles
[Zhao, Chaoyang]'s Articles
Baidu academic
Similar articles in Baidu academic
[Wang, Tong]'s Articles
[Zhu, Yousong]'s Articles
[Zhao, Chaoyang]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Wang, Tong]'s Articles
[Zhu, Yousong]'s Articles
[Zhao, Chaoyang]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: Wang2020_Chapter_LargeBatchOptimizationForObjec.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.