Pedestrian Attribute Recognition by Joint Visual-semantic Reasoning and Knowledge Distillation
Li, Qiaozhe; Zhao, Xin; He, Ran; Huang, Kaiqi
2019-08
会议名称28th International Joint Conference on Artificial Intelligence
会议日期2019-8
会议地点中国澳门
摘要

Pedestrian attribute recognition in surveillance is a challenging task in computer vision due to significant pose variation, viewpoint change and poor image quality. To achieve effective recognition, this paper presents a graph-based global reasoning framework to jointly model potential visual-semantic relations of attributes and distill auxiliary human parsing knowledge to guide the relational learning. The reasoning framework models attribute groups on a graph and learns a projection function to adaptively assign local visual features to the nodes of the graph. After feature projection, graph convolution is utilized to perform global reasoning between the attribute groups to model their mutual dependencies. Then, the learned node features are projected back to visual space to facilitate knowledge transfer. An additional regularization term is proposed by distilling human parsing knowledge from a pre-trained teacher model to enhance feature representations. The proposed framework is verified on three large scale pedestrian attribute datasets including PETA, RAP, and PA100k. Experiments show that our method achieves state-of-the-art results.

收录类别EI
七大方向——子方向分类图像视频处理与分析
文献类型会议论文
条目标识符http://ir.ia.ac.cn/handle/173211/28373
专题复杂系统认知与决策实验室_智能系统与工程
通讯作者Huang, Kaiqi
作者单位中国科学院自动化研究所
第一作者单位中国科学院自动化研究所
通讯作者单位中国科学院自动化研究所
推荐引用方式
GB/T 7714
Li, Qiaozhe,Zhao, Xin,He, Ran,et al. Pedestrian Attribute Recognition by Joint Visual-semantic Reasoning and Knowledge Distillation[C],2019.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
0117.pdf(640KB)会议论文 开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Li, Qiaozhe]的文章
[Zhao, Xin]的文章
[He, Ran]的文章
百度学术
百度学术中相似的文章
[Li, Qiaozhe]的文章
[Zhao, Xin]的文章
[He, Ran]的文章
必应学术
必应学术中相似的文章
[Li, Qiaozhe]的文章
[Zhao, Xin]的文章
[He, Ran]的文章
相关权益政策
暂无数据
收藏/分享
文件名: 0117.pdf
格式: Adobe PDF
此文件暂不支持浏览
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。