Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

contrastive estimators

Contrastive estimators are statistical objectives used in machine learning to estimate information-theoretic quantities, such as mutual information and probability density ratios, by training a model to distinguish between related data pairs and unrelated noise samples. Rather than directly calculating intractable probability distributions or normalization constants, these estimators formulate the estimation process as a discrimination task that scores paired data points drawn from a joint distribution higher than negative samples drawn from marginal distributions. By optimizing variational bounds on mutual information, contrastive estimators enable self-supervised and multimodal representation learning frameworks to capture meaningful shared dependencies and discard irrelevant variations across different views or modalities without requiring labeled data.

1 item

Factorized Contrastive Learning: Going Beyond Multi-view Redundancy

Factorized Contrastive Learning: Going Beyond Multi-view Redundancy

Paul Pu Liang, Zihao Deng, Martin Q. Ma, James Y. Zou, Louis-Philippe Morency, Ruslan Salakhutdinov

OrganizationsCarnegie Mellon UniversityStanford UniversityUniversity of Pennsylvania

Why you should read this

Proposes FACTORCL, a multimodal contrastive learning framework that overcomes standard multi-view redundancy limitations by factorizing representations into task-relevant shared and unique information via conditional mutual information bounds and multimodal augmentations.

In a wide range of multimodal tasks, contrastive learning has become a particularly appealing approach since it can successfully learn representations from abundant unlabeled data with only pairing information (e.g., image-caption or video-audio pairs). Underpinning these approaches is the assumption of multi-view redundancy - that shared information between modalities is necessary and sufficient for downstream tasks. However, in many real-world settings, task-relevant information is also contained in modality-unique regions: information that is only present in one modality but still relevant to the task. How can we learn self-supervised multimodal representations to capture both shared and unique information relevant to downstream tasks? This paper proposes FACTORCL, a new multimodal representation learning method to go beyond multi-view redundancy. FACTORCL is built from three new contributions: (1) factorizing task-relevant information into shared and unique representations, (2) capturing task-relevant information via maximizing MI lower bounds and removing task-irrelevant information via minimizing MI upper bounds, and (3) multimodal data augmentations to approximate task relevance without labels. On large-scale real-world datasets, FACTORCL captures both shared and unique information and achieves state-of-the-art results on six benchmarks.

Added

2026-09-26