中国科学院深圳先进技术研究院机构知识库(SIAT OpenIR): Bootstrap Model Ensemble and Rank Loss for Engagement Intensity Regression
SIAT OpenIR  > 数字所
Bootstrap Model Ensemble and Rank Loss for Engagement Intensity Regression
KaiWang; Jianfei Yang; Da Guo; Kaipeng Zhang; Xiaojiang Peng; Yu Qiao
2019
Conference NameICMI ’19
Conference Date2019
Conference PlaceBoulder, CO, USA
AbstractThis paper presents our approach for group-level emotion recogni-tion sub-challenge in the EmotiW 2018. The task is to classify animage into one of the group emotions such as positive, negative,and neutral. Our approach mainly exploits three types of visualcues for this task, namely face, body and global image with recentdeep networks. Our main contribution is two-fold. First, we intro-duce body Convolutional Neural Networks (CNNs) into this taskbased on our previous winner method [18]. Specially, we crop allbodies in an image with the state-of-the-art human pose estima-tion method and train body CNNs with the image-level labels ofgroup emotions. The body cue captures a full view of an individual.Second, we propose a cascade attention network for the face cue
Department多媒体
Indexed ByEI
Language英语
Document Type会议论文
Identifierhttp://ir.siat.ac.cn/handle/172644/15909
Collection数字所
Recommended Citation
GB/T 7714
KaiWang,Jianfei Yang,Da Guo,et al. Bootstrap Model Ensemble and Rank Loss for Engagement Intensity Regression[C],2019.
Files in This Item: Download All
File Name/Size DocType Version Access License
数字-多媒体2019025.pdf(957KB)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
[KaiWang]'s Articles
[Jianfei Yang]'s Articles
[Da Guo]'s Articles
Baidu academic
Similar articles in Baidu academic
[KaiWang]'s Articles
[Jianfei Yang]'s Articles
[Da Guo]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[KaiWang]'s Articles
[Jianfei Yang]'s Articles
[Da Guo]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 数字-多媒体2019025.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

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