Proceedings of the 5th International Conference on Multimedia and Image Processing最新文献

筛选

英文中文

3DoF+ video compression algorithm based on distortion minimization interpolation with adaptive reference view selection 基于失真最小化插值和自适应参考视图选择的3DoF+视频压缩算法

Proceedings of the 5th International Conference on Multimedia and Image Processing Pub Date : 2020-01-10 DOI: 10.1145/3381271.3381279

Mingliu Sun, Shengyang Zhao, Xiaoming Chen

{"title":"3DoF+ video compression algorithm based on distortion minimization interpolation with adaptive reference view selection","authors":"Mingliu Sun, Shengyang Zhao, Xiaoming Chen","doi":"10.1145/3381271.3381279","DOIUrl":"https://doi.org/10.1145/3381271.3381279","url":null,"abstract":"The emerging 3DoF+ video offers more interactivity and deeper immersion. However, the bulky size of the 3DoF+ video imposes new challenges in its storage and processing. Thus, compressing 3DoF+ video efficiently is vitally important. Due to the geometric relationship between viewpoints, overlaps exist in views and this redundancy is available for further compression. However, existing compression algorithms either utilize partial information of the scene or still retain a lot of redundancy, which results in low quality reconstructed views. To address these problems, our study proposes to select reference views adaptively, which utilizes as few bits as possible to contain the necessary visual information of the scene. Besides, in order to reduce the residues between the reconstructed and original views, the distortion minimization interpolation (DMI) method is then proposed for further refining the synthesized views. The experimental results show that our method achieves on average 52.1% BD-rate reduction than direct coding with High Efficiency Video Coding (HEVC).","PeriodicalId":124651,"journal":{"name":"Proceedings of the 5th International Conference on Multimedia and Image Processing","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122418727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Two-stage metric learning for cross-modality person re-identification 跨模态人再识别的两阶段度量学习

Proceedings of the 5th International Conference on Multimedia and Image Processing Pub Date : 2020-01-10 DOI: 10.1145/3381271.3381285

Jiabao Wang, Shanshan Jiao, Yang Li, Zhuang Miao

引用次数: 5

首页上一页