RPN+ fast boosted tree: Combining deep neural network with traditional classifier for pedestrian detection

Jiaxiang Zhao, Jun Li, Yingdong Ma
{"title":"RPN+ fast boosted tree: Combining deep neural network with traditional classifier for pedestrian detection","authors":"Jiaxiang Zhao, Jun Li, Yingdong Ma","doi":"10.1109/CATA.2018.8398672","DOIUrl":null,"url":null,"abstract":"The problem of pedestrian detection receives increasing attention due to the rapid development of artificial intelligence technologies. In this paper, we propose a deep neural network based method which combines with a traditional classifier for fast and robust pedestrian detection. Specifically, region proposals generation and feature extraction are implemented using a modified RPN-VGG method. The proposed method is designed to improve system performance on small objects detection. A new classifier, Fast Boosted Tree, is trained based on RPN outputs to obtain the final results. Experiments on Caltech pedestrian dataset demonstrate that the proposed method achieves 8.77% miss rate and has the best known efficiency with state-of-the-art CNN-based detectors. When algorithm efficiency is not considered, detection quality can be further improved to 8.25% miss rate by adding global normalization and optical flow features.","PeriodicalId":231024,"journal":{"name":"2018 4th International Conference on Computer and Technology Applications (ICCTA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"8","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 4th International Conference on Computer and Technology Applications (ICCTA)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CATA.2018.8398672","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 8

Abstract

The problem of pedestrian detection receives increasing attention due to the rapid development of artificial intelligence technologies. In this paper, we propose a deep neural network based method which combines with a traditional classifier for fast and robust pedestrian detection. Specifically, region proposals generation and feature extraction are implemented using a modified RPN-VGG method. The proposed method is designed to improve system performance on small objects detection. A new classifier, Fast Boosted Tree, is trained based on RPN outputs to obtain the final results. Experiments on Caltech pedestrian dataset demonstrate that the proposed method achieves 8.77% miss rate and has the best known efficiency with state-of-the-art CNN-based detectors. When algorithm efficiency is not considered, detection quality can be further improved to 8.25% miss rate by adding global normalization and optical flow features.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
RPN+快速提升树:将深度神经网络与传统分类器相结合用于行人检测
随着人工智能技术的飞速发展,行人检测问题越来越受到人们的关注。本文提出了一种基于深度神经网络和传统分类器相结合的快速鲁棒行人检测方法。具体来说,使用改进的RPN-VGG方法实现区域建议生成和特征提取。该方法旨在提高系统在小目标检测方面的性能。在RPN输出的基础上训练一种新的分类器,快速提升树,以获得最终结果。在加州理工学院行人数据集上的实验表明,该方法的缺失率为8.77%,是目前最先进的基于cnn的检测器中效率最高的。在不考虑算法效率的情况下,通过加入全局归一化和光流特征,可以将检测质量进一步提高到8.25%的缺失率。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Classification of pornographic content on Twitter using support vector machine and Naive Bayes Concepts extraction in ontology learning using language patterns for better accuracy Automatically generate a specific human computer interaction from an interface diagram model The Triple Helix Model: University-industry-governments linkage web-based application recommendation systems for emerging commercial-base research State of the art of telepresence with a virtual reality headset
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1