AmCLR: Unified Augmented Learning for Cross-Modal Representations

TL;DR

AmCLR and xAmCLR are novel contrastive learning frameworks for vision-language models that improve robustness and efficiency by integrating diverse augmentations and intra-modal alignments, reducing computational requirements.

Contribution

The paper introduces AmCLR and xAmCLR, new contrastive learning objectives that enhance bimodal representation learning with fewer resources and richer augmentations.

Findings

AmCLR achieves comparable performance with smaller batch sizes.

xAmCLR improves intra-modal alignment for better feature richness.

Both methods demonstrate increased robustness and efficiency.

Abstract

Contrastive learning has emerged as a pivotal framework for representation learning, underpinning advances in both unimodal and bimodal applications like SimCLR and CLIP. To address fundamental limitations like large batch size dependency and bimodality, methods such as SogCLR leverage stochastic optimization for the global contrastive objective. Inspired by SogCLR's efficiency and adaptability, we introduce AmCLR and xAmCLR objective functions tailored for bimodal vision-language models to further enhance the robustness of contrastive learning. AmCLR integrates diverse augmentations, including text paraphrasing and image transformations, to reinforce the alignment of contrastive representations, keeping batch size limited to a few hundred samples unlike CLIP which needs batch size of 32,768 to produce reasonable results. xAmCLR further extends this paradigm by incorporating intra-modal alignments between original and augmented modalities for richer feature learning. These advancements yield a more resilient and generalizable contrastive learning process, aimed at overcoming bottlenecks in scaling and augmentative diversity. Since we have built our framework on the existing SogCLR, we are able to demonstrate improved representation quality with fewer computational resources, establishing a foundation for scalable and robust multi-modal learning.