XDBERT: Distilling Visual Information to BERT from Cross-Modal Systems to Improve Language Understanding
Chan-Jan Hsu, Hung-yi Lee, Yu Tsao
Abstract
Transformer-based models are widely used in natural language understanding (NLU) tasks, and multimodal transformers have been effective in visual-language tasks. This study explores distilling visual information from pretrained multimodal transformers to pretrained language encoders. Our framework is inspired by cross-modal encoders’ success in visual-language tasks while we alter the learning objective to cater to the language-heavy characteristics of NLU. After training with a small number of extra adapting steps and finetuned, the proposed XDBERT (cross-modal distilled BERT) outperforms pretrained-BERT in general language understanding evaluation (GLUE), situations with adversarial generations (SWAG) benchmarks, and readability benchmarks. We analyze the performance of XDBERT on GLUE to show that the improvement is likely visually grounded.
BibTeX
@inproceedings{hsu-etal-2022-xdbert,
title = "{XDBERT}: {D}istilling Visual Information to {BERT} from Cross-Modal Systems to Improve Language Understanding",
author = "Hsu, Chan-Jan and
Lee, Hung-yi and
Tsao, Yu",
editor = "Muresan, Smaranda and
Nakov, Preslav and
Villavicencio, Aline",
booktitle = "Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = may,
year = "2022",
address = "Dublin, Ireland",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.acl-short.52/",
doi = "10.18653/v1/2022.acl-short.52",
pages = "479--489"
}