Differentiable Dynamics for Articulated 3d Human Motion Reconstruction

Erik Gärtner, Mykhaylo Andriluka, Erwin Coumans, Cristian Sminchisescu

Forskningsoutput: Kapitel i bok/rapport/Conference proceedingKonferenspaper i proceedingPeer review

Sammanfattning

We introduce DiffPhy, a differentiable physics-based model for articulated 3d human motion reconstruction from video. Applications of physics-based reasoning in human motion analysis have so far been limited, both by the complexity of constructing adequate physical models of articulated human motion, and by the formidable challenges of performing stable and efficient inference with physics in the loop. We jointly address such modeling and inference challenges by proposing an approach that combines a physically plausible body representation with anatomical joint limits, a differentiable physics simulator, and optimization techniques that ensure good performance and robustness to suboptimal local optima. In contrast to several recent methods [39], [42], [55], our approach readily supports full-body contact including interactions with objects in the scene. Most importantly, our model connects end-to-end with images, thus supporting direct gradient-based physics optimization by means of image-based loss functions. We validate the model by demonstrating that it can accurately reconstruct physically plausible 3d human motion from monocular video, both on public benchmarks with available 3d ground-truth, and on videos from the internet.
Originalspråkengelska
Titel på värdpublikationProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition
FörlagIEEE - Institute of Electrical and Electronics Engineers Inc.
ISBN (elektroniskt)978-1-6654-6946-3
ISBN (tryckt)978-1-6654-6947-0
DOI
StatusPublished - 2022
Evenemang2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022 - New Orleans, USA
Varaktighet: 2022 juni 192022 juni 24

Konferens

Konferens2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022
Land/TerritoriumUSA
OrtNew Orleans
Period2022/06/192022/06/24

Ämnesklassifikation (UKÄ)

  • Datorseende och robotik (autonoma system)

Fingeravtryck

Utforska forskningsämnen för ”Differentiable Dynamics for Articulated 3d Human Motion Reconstruction”. Tillsammans bildar de ett unikt fingeravtryck.

Citera det här