File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1007/978-3-031-20065-6_35
- Scopus: eid_2-s2.0-85144490710
- Find via
Supplementary
-
Citations:
- Scopus: 0
- Appears in Collections:
Conference Paper: DeciWatch: A Simple Baseline for 10 × Efficient 2D and 3D Pose Estimation
Title | DeciWatch: A Simple Baseline for 10 × Efficient 2D and 3D Pose Estimation |
---|---|
Authors | |
Keywords | Efficiency Human Pose Estimation Video Analysis |
Issue Date | 2022 |
Citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, v. 13665 LNCS, p. 607-624 How to Cite? |
Abstract | This paper proposes a simple baseline framework for video-based 2D/3D human pose estimation that can achieve 10 × efficiency improvement over existing works without any performance degradation, named DeciWatch. Unlike current solutions that estimate each frame in a video, DeciWatch introduces a simple yet effective sample-denoise-recover framework that only watches sparsely sampled frames, taking advantage of the continuity of human motions and the lightweight pose representation. Specifically, DeciWatch uniformly samples less than 10 % video frames for detailed estimation, denoises the estimated 2D/3D poses with an efficient Transformer architecture, and then accurately recovers the rest of the frames using another Transformer-based network. Comprehensive experimental results on three video-based human pose estimation, body mesh recovery tasks and efficient labeling in videos with four datasets validate the efficiency and effectiveness of DeciWatch. Code is available at https://github.com/cure-lab/DeciWatch. |
Persistent Identifier | http://hdl.handle.net/10722/352336 |
ISSN | 2023 SCImago Journal Rankings: 0.606 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Zeng, Ailing | - |
dc.contributor.author | Ju, Xuan | - |
dc.contributor.author | Yang, Lei | - |
dc.contributor.author | Gao, Ruiyuan | - |
dc.contributor.author | Zhu, Xizhou | - |
dc.contributor.author | Dai, Bo | - |
dc.contributor.author | Xu, Qiang | - |
dc.date.accessioned | 2024-12-16T03:58:20Z | - |
dc.date.available | 2024-12-16T03:58:20Z | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, v. 13665 LNCS, p. 607-624 | - |
dc.identifier.issn | 0302-9743 | - |
dc.identifier.uri | http://hdl.handle.net/10722/352336 | - |
dc.description.abstract | This paper proposes a simple baseline framework for video-based 2D/3D human pose estimation that can achieve 10 × efficiency improvement over existing works without any performance degradation, named DeciWatch. Unlike current solutions that estimate each frame in a video, DeciWatch introduces a simple yet effective sample-denoise-recover framework that only watches sparsely sampled frames, taking advantage of the continuity of human motions and the lightweight pose representation. Specifically, DeciWatch uniformly samples less than 10 % video frames for detailed estimation, denoises the estimated 2D/3D poses with an efficient Transformer architecture, and then accurately recovers the rest of the frames using another Transformer-based network. Comprehensive experimental results on three video-based human pose estimation, body mesh recovery tasks and efficient labeling in videos with four datasets validate the efficiency and effectiveness of DeciWatch. Code is available at https://github.com/cure-lab/DeciWatch. | - |
dc.language | eng | - |
dc.relation.ispartof | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | - |
dc.subject | Efficiency | - |
dc.subject | Human Pose Estimation | - |
dc.subject | Video Analysis | - |
dc.title | DeciWatch: A Simple Baseline for 10 × Efficient 2D and 3D Pose Estimation | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1007/978-3-031-20065-6_35 | - |
dc.identifier.scopus | eid_2-s2.0-85144490710 | - |
dc.identifier.volume | 13665 LNCS | - |
dc.identifier.spage | 607 | - |
dc.identifier.epage | 624 | - |
dc.identifier.eissn | 1611-3349 | - |