Naslov
Generalizable Human Pose Triangulation

Autori
Bartol, Kristijan ; Bojanić, David ; Petković, Tomislav ; Pribanić, Tomislav

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) / Chellapa, Rama ; Matas, Jiri ; Quan, Long ; Shah, Mubarak - New Orleans (LA) : Institute of Electrical and Electronics Engineers (IEEE), 2022, 11018-11027

ISBN
978-1-6654-6946-3

Skup
2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Mjesto i datum
New Orleans (LA), Sjedinjene Američke Države, 18.06.2022. - 24.06.2022

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
human pose ; stochastic learning ; generalizable triangulation

Sažetak
We address the problem of generalizability for multi-view 3D human pose estimation. The standard approach is to first detect 2D keypoints in images and then apply triangulation from multiple views. Even though the existing methods achieve remarkably accurate 3D pose estimation on public benchmarks, most of them are limited to a single spatial camera arrangement and their number. Several methods address this limitation but demonstrate significantly degraded performance on novel views. We propose a stochastic framework for human pose triangulation and demonstrate a superior generalization across different camera arrangements on two public datasets. In addition, we apply the same approach to the fundamental matrix estimation problem, showing that the proposed method can successfully apply to other computer vision problems. The stochastic framework achieves more than 8.8% improvement on the 3D pose estimation task, compared to the state-of-the-art, and more than 30% improvement for fundamental matrix estimation, compared to a standard algorithm.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo

POVEZANOST RADA