CLAP: Isolating Content from Style through Contrastive Learning with Augmented Prompts
This work addresses a key limitation in multimodal learning for researchers and practitioners by refining vision-language representations, though it is incremental as it builds on existing CLIP-like models.
The paper tackled the problem of contrastive vision-language models blending content and style information, which limits generalization under distribution shifts, by proposing CLAP to isolate content from style through contrastive learning with augmented prompts, resulting in significant improvements in zero-shot and few-shot classification tasks and enhanced robustness to perturbations.
Contrastive vision-language models, such as CLIP, have garnered considerable attention for various downstream tasks, mainly due to the remarkable ability of the learned features for generalization. However, the features they learned often blend content and style information, which somewhat limits their generalization capabilities under distribution shifts. To address this limitation, we adopt a causal generative perspective for multimodal data and propose contrastive learning with data augmentation to disentangle content features from the original representations. To achieve this, we begin with exploring image augmentation techniques and develop a method to seamlessly integrate them into pre-trained CLIP-like models to extract pure content features. Taking a step further, recognizing the inherent semantic richness and logical structure of text data, we explore the use of text augmentation to isolate latent content from style features. This enables CLIP-like model's encoders to concentrate on latent content information, refining the learned representations by pre-trained CLIP-like models. Our extensive experiments across diverse datasets demonstrate significant improvements in zero-shot and few-shot classification tasks, alongside enhanced robustness to various perturbations. These results underscore the effectiveness of our proposed methods in refining vision-language representations and advancing the state-of-the-art in multimodal learning.