ALIP: Adaptive Language-Image Pre-training with Synthetic Caption

TL;DR

ALIP introduces an adaptive pre-training method that utilizes synthetic captions and dynamic weighting mechanisms to improve vision-language representations, reducing noise impact and enhancing downstream task performance.

Contribution

This work proposes ALIP, a novel adaptive pre-training framework that integrates synthetic captions and dynamic sample weighting to improve vision-language models.

Findings

Achieves state-of-the-art results on zero-shot image-text retrieval.

Effectively reduces noise impact during pre-training.

Enhances downstream task performance across various models and datasets.

Abstract

Contrastive Language-Image Pre-training (CLIP) has significantly boosted the performance of various vision-language tasks by scaling up the dataset with image-text pairs collected from the web. However, the presence of intrinsic noise and unmatched image-text pairs in web data can potentially affect the performance of representation learning. To address this issue, we first utilize the OFA model to generate synthetic captions that focus on the image content. The generated captions contain complementary information that is beneficial for pre-training. Then, we propose an Adaptive Language-Image Pre-training (ALIP), a bi-path model that integrates supervision from both raw text and synthetic caption. As the core components of ALIP, the Language Consistency Gate (LCG) and Description Consistency Gate (DCG) dynamically adjust the weights of samples and image-text/caption pairs during the training process. Meanwhile, the adaptive contrastive loss can effectively reduce the impact of noise data and enhances the efficiency of pre-training data. We validate ALIP with experiments on different scales of models and pre-training datasets. Experiments results show that ALIP achieves state-of-the-art performance on multiple downstream tasks including zero-shot image-text retrieval and linear probe. To facilitate future research, the code and pre-trained models are released at https://github.com/deepglint/ALIP.