Learning interpretable representation for 3D point clouds
Journal
Proceedings - International Conference on Pattern Recognition
Pages
7470-7477
Date Issued
2020
Author(s)
Abstract
Point clouds have emerged as a popular representation of 3D visual data. With a set of unordered 3D points, one typically needs to transform them into latent representation before further classification and segmentation tasks. However, one cannot easily interpret such encoded latent representation. To address this issue, we propose a unique deep learning framework for disentangling body-type and pose information from 3D point clouds. Extending from autoencoder, we advance adversarial learning a selected feature type, while classification and data recovery can be additionally observed. Our experiments confirm that our model can be successfully applied to perform a wide range of 3D applications like shape synthesis, action translation, shape/action interpolation, and synchronization. © 2021 IEEE
Other Subjects
Deep learning; Pattern recognition; 3D application; 3D point cloud; 3D visual data; Adversarial learning; Interpretable representation; Learning frameworks; Pose information; Shape synthesis; Three dimensional computer graphics
Type
conference paper