VLMo: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts
Hangbo Bao, Wenhui Wang, Li Dong, Qiang Liu, Owais Khan Mohammed, Kriti Aggarwal, Subhojit Som, Songhao Piao
Abstract
We present a unified Vision-Language pretrained Model (VLMo) that jointly learns a dual encoder and a fusion encoder with a modular Transformer network. Specifically, we introduce Multiway Transformer, where each block contains a pool of modality-specific experts and a shared self-attention layer. Because of the modeling flexibility of Multiway Transformer, pretrained VLMo can be fine-tuned as a fusion encoder for vision-language classification tasks, or used as a dual encoder for efficient image-text retrieval. Moreover, we propose a stagewise pre-training strategy, which effectively leverages large-scale image-only and text-only data besides image-text pairs. Experimental results show that VLMo achieves state-of-the-art results on various vision-language tasks, including VQA, NLVR2 and image-text retrieval.
BibTeX
@inproceedings{
bao2022vlmo,
title={{VLM}o: Unified Vision-Language Pre-Training with Mixture-of-Modality-Experts},
author={Hangbo Bao and Wenhui Wang and Li Dong and Qiang Liu and Owais Khan Mohammed and Kriti Aggarwal and Subhojit Som and Songhao Piao and Furu Wei},
booktitle={Advances in Neural Information Processing Systems},
editor={Alice H. Oh and Alekh Agarwal and Danielle Belgrave and Kyunghyun Cho},
year={2022},
url={https://openreview.net/forum?id=bydKs84JEyw}
}