Reconstructing Human Body Mesh from Point Clouds by Adversarial GP Network - Archive ouverte HAL Access content directly
Conference Papers Year : 2021

Reconstructing Human Body Mesh from Point Clouds by Adversarial GP Network

(1, 2) , (1) , (3) , (3) , (1)
1
2
3
Federica Bogo
  • Function : Author
  • PersonId : 1080019
Bugra Tekin
  • Function : Author
  • PersonId : 1080020
Edmond Boyer

Abstract

We study the problem of reconstructing the template-aligned mesh for human body estimation from unstructured point cloud data. Recently proposed approaches for shape matching that rely on Deep Neural Networks (DNNs) achieve state-of-the-art results with generic point-wise architectures; but in doing so, they exploit much weaker human body shape and surface priors with respect to methods that explicitly model the body surface with 3D templates. We investigate the impact of adding back such stronger shape priors by proposing a novel dedicated human template matching process, which relies on a point-based, deep autoencoder architecture. We encode surface smoothness and shape coherence with a specialized Gaussian Process layer. Furthermore, we enforce global consistency and improve the generalization capabilities of the model by introducing an adversarial training phase. The choice of these elements is grounded on an extensive analysis of DNNs failure modes in widely used datasets like SURREAL and FAUST. We validate and evaluate the impact of our novel components on these datasets, showing a quantitative improvement over state-of-the-art DNN-based methods, and qualitatively better results.
Fichier principal
Vignette du fichier
accv.pdf (3.32 Mo) Télécharger le fichier
Vignette du fichier
accv-supp.pdf (3.92 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Origin : Files produced by the author(s)

Dates and versions

hal-02977388 , version 1 (24-10-2020)

Identifiers

Cite

Boyao Zhou, Jean-Sébastien Franco, Federica Bogo, Bugra Tekin, Edmond Boyer. Reconstructing Human Body Mesh from Point Clouds by Adversarial GP Network. ACCV - Asian Conference on Computer Vision, Nov 2020, Kyoto, Japan. pp.123-139, ⟨10.1007/978-3-030-69525-5_8⟩. ⟨hal-02977388⟩
235 View
767 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More