hh.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
3D human pose and shape estimation via de-occlusion multi-task learning
Chinese Academy of Sciences, Beijing, China; University of Chinese Academy of Sciences, Beijing, China.
Chinese Academy of Sciences, Beijing, China; University of Chinese Academy of Sciences, Beijing, China; Cognitive Computing Technology Joint Laboratory, Wave Group, Beijing, China.ORCID iD: 0000-0001-7897-1673
Chinese Academy of Sciences, Beijing, China; University of Chinese Academy of Sciences, Beijing, China; Beijing Key Laboratory Of Semiconductor Neural Network Intelligent Sensing and Computing Technology, Beijing, China.ORCID iD: 0000-0001-9668-2883
Chinese Academy of Sciences, Beijing, China; Hebei University of Engineering, Handan, China.
Show others and affiliations
2023 (English)In: Neurocomputing, ISSN 0925-2312, E-ISSN 1872-8286, Vol. 548, article id 126284Article in journal (Refereed) Published
Abstract [en]

Three-dimensional human pose and shape estimation is to compute a full human 3D mesh given a single image. The contamination of features caused by occlusion usually degrades its performance significantly. Recent progress in this field typically addressed the occlusion problem implicitly. By contrast, in this paper, we address it explicitly using a simple yet effective de-occlusion multi-task learning network. Our key insight is that feature for mesh parameter regression should be noiseless. Thus, in the feature space, our method disentangles the occludee that represents the noiseless human feature from the occluder. Specifically, a spatial regularization and an attention mechanism are imposed in the backbone of our network to disentangle the features into different channels. Furthermore, two segmentation tasks are proposed to supervise the de-occlusion process. The final mesh model is regressed by the disentangled occlusion-aware features. Experiments on both occlusion and non-occlusion datasets are conducted, and the results prove that our method is superior to the state-of-the-art methods on two occlusion datasets, while achieving competitive performance on a non-occlusion dataset. We also demonstrate that the proposed de-occlusion strategy is the main factor to improve the robustness against occlusion. The code is available at https://github.com/qihangran/De-occlusion_MTL_HMR. © 2023

Place, publisher, year, edition, pages
Amsterdam: Elsevier, 2023. Vol. 548, article id 126284
Keywords [en]
De-occlusion, Human mesh recovery, Multi-task learning, Occlusion-aware
National Category
Computer graphics and computer vision
Identifiers
URN: urn:nbn:se:hh:diva-51446DOI: 10.1016/j.neucom.2023.126284ISI: 001025622900001Scopus ID: 2-s2.0-85161577595OAI: oai:DiVA.org:hh-51446DiVA, id: diva2:1789021
Note

Funding: This work is supported by the National Natural Science Foundation of China (Grant No. 61901436).

Available from: 2023-08-17 Created: 2023-08-17 Last updated: 2025-02-07Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textScopus

Authority records

Tiwari, Prayag

Search in DiVA

By author/editor
Ning, XinLi, WeijunTiwari, Prayag
By organisation
School of Information Technology
In the same journal
Neurocomputing
Computer graphics and computer vision

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 98 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf