ELIP: Efficient Discriminative Language-Image Pre-training with Fewer Vision Tokens

TL;DR

ELIP introduces a vision token pruning method for language-image pre-training that reduces computational costs while maintaining performance, enabling faster training and larger batch sizes.

Contribution

The paper presents a novel, task-aligned token pruning approach for efficient language-image pre-training that is computation- and memory-efficient, and parameter-free.

Findings

Removes ~30% vision tokens with minimal performance loss (~0.32 accuracy drop)

Maintains performance across various downstream tasks

Enables larger batch sizes and faster pre-training

Abstract

Learning a versatile language-image model is computationally prohibitive under a limited computing budget. This paper delves into the \emph{efficient language-image pre-training}, an area that has received relatively little attention despite its importance in reducing computational cost and footprint. To that end, we propose a vision token pruning and merging method ELIP, to remove less influential tokens based on the supervision of language outputs. Our method is designed with several strengths, such as being computation-efficient, memory-efficient, and trainable-parameter-free, and is distinguished from previous vision-only token pruning approaches by its alignment with task objectives. We implement this method in a progressively pruning manner using several sequential blocks. To evaluate its generalization performance, we apply ELIP to three commonly used language-image pre-training models and utilize public image-caption pairs with 4M images for pre-training. Our experiments demonstrate that with the removal of ~30$\%$ vision tokens across 12 ViT layers, ELIP maintains significantly comparable performance with baselines ($\sim$0.32 accuracy drop on average) over various downstream tasks including cross-modal retrieval, VQA, image captioning, \emph{etc}. In addition, the spared GPU resources by our ELIP allow us to scale up with larger batch sizes, thereby accelerating model pre-training and even sometimes enhancing downstream model performance.