Generative Zero-Shot Learning for Semantic Segmentation of 3D Point Clouds - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Generative Zero-Shot Learning for Semantic Segmentation of 3D Point Clouds

Bjorn Michele
  • Fonction : Auteur
Alexandre Boulch
  • Fonction : Auteur
Gilles Puy
  • Fonction : Auteur
Maxime Bucher
  • Fonction : Auteur

Résumé

While there has been a number of studies on Zero-Shot Learning (ZSL) for 2D images, its application to 3D data is still recent and scarce, with just a few methods limited to classification. We present the first generative approach for both ZSL and Generalized ZSL (GZSL) on 3D data, that can handle both classification and, for the first time, semantic segmentation. We show that it reaches or outperforms the state of the art on ModelNet40 classification for both inductive ZSL and inductive GZSL. For semantic segmentation, we created three benchmarks for evaluating this new ZSL task, using S3DIS, ScanNet and SemanticKITTI. Our experiments show that our method outperforms strong baselines, which we additionally propose for this task.

Dates et versions

hal-04489208 , version 1 (04-03-2024)

Identifiants

Citer

Bjorn Michele, Alexandre Boulch, Gilles Puy, Maxime Bucher, Renaud Marlet. Generative Zero-Shot Learning for Semantic Segmentation of 3D Point Clouds. 3DV 2021, Dec 2021, London, United Kingdom. pp.992-1002, ⟨10.1109/3DV53792.2021.00107⟩. ⟨hal-04489208⟩
5 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More