Info3D: Representation Learning on 3D Objects using Mutual Information Maximization and Contrastive Learning

TL;DR

This paper introduces Info3D, a novel unsupervised learning framework for 3D shape analysis that maximizes mutual information and contrastive learning to improve representation quality, invariance, and downstream task performance.

Contribution

It extends mutual information and contrastive learning principles to 3D shapes, enabling rotation invariance and better representations without requiring aligned or reconstructed datasets.

Findings

Achieves state-of-the-art results in shape clustering and retrieval.

Demonstrates effective rotation invariance in 3D shape representations.

Improves transfer learning performance on 3D shape tasks.

Abstract

A major endeavor of computer vision is to represent, understand and extract structure from 3D data. Towards this goal, unsupervised learning is a powerful and necessary tool. Most current unsupervised methods for 3D shape analysis use datasets that are aligned, require objects to be reconstructed and suffer from deteriorated performance on downstream tasks. To solve these issues, we propose to extend the InfoMax and contrastive learning principles on 3D shapes. We show that we can maximize the mutual information between 3D objects and their "chunks" to improve the representations in aligned datasets. Furthermore, we can achieve rotation invariance in SO${(3)}$ group by maximizing the mutual information between the 3D objects and their geometric transformed versions. Finally, we conduct several experiments such as clustering, transfer learning, shape retrieval, and achieve state of art results.