LiteVL: Efficient Video-Language Learning with Enhanced Spatial-Temporal Modeling

TL;DR

LiteVL is a computationally efficient video-language model that adapts a pre-trained image-language model for video tasks by adding temporal attention and adaptive pooling, achieving superior results without heavy pre-training.

Contribution

It introduces a novel adaptation method for image-language models to handle video tasks efficiently, eliminating the need for extensive pre-training.

Findings

Outperforms previous models on text-video retrieval

Achieves better results on video question answering

Operates without heavy video-language pre-training

Abstract

Recent large-scale video-language pre-trained models have shown appealing performance on various downstream tasks. However, the pre-training process is computationally expensive due to the requirement of millions of video-text pairs and the redundant data structure of each video. To mitigate these problems, we propose LiteVL, which adapts a pre-trained image-language model BLIP into a video-text model directly on downstream tasks, without heavy pre-training. To enhance the temporal modeling lacking in the image-language model, we propose to add temporal attention modules in the image encoder of BLIP with dynamic temporal scaling. Besides the model-wise adaptation, we also propose a non-parametric pooling mechanism to adaptively reweight the fine-grained video embedding conditioned on the text. Experimental results on text-video retrieval and video question answering show that the proposed LiteVL even outperforms previous video-language pre-trained models by a clear margin, though without any video-language pre-training.