Random Walks in Self-supervised Learning for Triangular Meshes

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yefet, Gal, Tal, Ayellet
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910853907873792
author Yefet, Gal
Tal, Ayellet
author_facet Yefet, Gal
Tal, Ayellet
contents This study addresses the challenge of self-supervised learning for 3D mesh analysis. It presents an new approach that uses random walks as a form of data augmentation to generate diverse representations of mesh surfaces. Furthermore, it employs a combination of contrastive and clustering losses. The contrastive learning framework maximizes similarity between augmented instances of the same mesh while minimizing similarity between different meshes. We integrate this with a clustering loss, enhancing class distinction across training epochs and mitigating training variance. Our model's effectiveness is evaluated using mean Average Precision (mAP) scores and a supervised SVM linear classifier on extracted features, demonstrating its potential for various downstream tasks such as object classification and shape retrieval.
format Preprint
id arxiv_https___arxiv_org_abs_2503_00816
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Random Walks in Self-supervised Learning for Triangular Meshes
Yefet, Gal
Tal, Ayellet
Graphics
Computer Vision and Pattern Recognition
Machine Learning
This study addresses the challenge of self-supervised learning for 3D mesh analysis. It presents an new approach that uses random walks as a form of data augmentation to generate diverse representations of mesh surfaces. Furthermore, it employs a combination of contrastive and clustering losses. The contrastive learning framework maximizes similarity between augmented instances of the same mesh while minimizing similarity between different meshes. We integrate this with a clustering loss, enhancing class distinction across training epochs and mitigating training variance. Our model's effectiveness is evaluated using mean Average Precision (mAP) scores and a supervised SVM linear classifier on extracted features, demonstrating its potential for various downstream tasks such as object classification and shape retrieval.
title Random Walks in Self-supervised Learning for Triangular Meshes
topic Graphics
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2503.00816