NEVLP: Noise-Robust Framework for Efficient Vision-Language Pre-training

TL;DR

NEVLP introduces a noise-robust, efficient pre-training framework for vision-language models that reduces data requirements and improves performance by mitigating noise through innovative learning strategies.

Contribution

The paper proposes NEVLP, a novel framework with noise-adaptive and concept-enhanced learning strategies, enabling effective pre-training with less data and noise robustness.

Findings

Achieves state-of-the-art results on multiple vision-language tasks.

Reduces the need for large-scale web data in pre-training.

Effectively mitigates noise impact in web-crawled datasets.

Abstract

The success of Vision Language Models (VLMs) on various vision-language tasks heavily relies on pre-training with large scale web-crawled datasets. However, the noisy and incomplete nature of web data makes dataset scale crucial for performance, rendering end-to-end training increasingly prohibitive. In this paper, we propose NEVLP, a noise-robust framework for efficient vision-language pre-training that requires less pre-training data. Specifically, we bridge the modality gap between a frozen image encoder and a large language model with a transformer and introduce two innovative learning strategies: noise-adaptive learning and concept-enhanced learning to mitigate the impact of noise. In noise-adaptive learning, we estimate the noise probability of each image-text pair based on the transformer's memorization effect and employ noise-adaptive regularization on image-text contrastive learning to condition cross-modal alignment. In concept-enhanced learning, we enrich incomplete text by incorporating visual concepts (objects in the image) to provide prior information about existing objects for image-text matching and image-grounded text generation, thereby mitigating text incompletion. Our framework effectively utilizes noisy web data and achieves state-of-the-art performance with less pre-training data across a wide range of vision-language tasks, including image-text retrieval, image captioning, and visual question answering.