Share this page:

An Empirical Study of Training End-to-End Vision-and-Language Transformers

Zi-Yi Dou, Yichong Xu, Zhe Gan, Jianfeng Wang, Shuohang Wang, Lijuan Wang, Chenguang Zhu, Pengchuan Zhang, Lu Yuan, Nanyun Peng, Zicheng Liu, and Michael Zeng, in The Conference on Computer Vision and Pattern Recognition (CVPR-22), 2022.

Download the full text


Abstract

Vision-and-language (VL) pre-training has proven to be highly effective on various VL downstream tasks. While recent work has shown that fully transformer-based VL models can be more efficient than previous region-feature-based methods, their performance on downstream tasks often degrades significantly. In this paper, we present METER, a Multimodal End-to-end TransformER framework, through which we investigate how to design and pre-train a fully transformer-based VL model in an end-to-end manner. Specifically, we dissect the model designs along multiple dimensions: vision encoders (e.g., CLIP-ViT, Swin transformer), text encoders (e.g., RoBERTa, DeBERTa), multimodal fusion module (e.g., merged attention vs. co-attention), architectural design (e.g., encoder-only vs. encoder-decoder), and pre-training objectives (e.g., masked image modeling). We conduct comprehensive experiments and provide insights on how to train a performant VL transformer while maintaining fast inference speed. Notably, our best model achieves an accuracy of 77.64% on the VQAv2 test-std set using only 4M images for pre-training, surpassing the state-of-the-art region-feature-based model by 1.04%, and outperforming the previous best fully transformer-based model by 1.6%.


Bib Entry

@inproceedings{dou2022meter,
  title = {An Empirical Study of Training End-to-End Vision-and-Language Transformers},
  author = {Dou, Zi-Yi and Xu, Yichong and Gan, Zhe and Wang, Jianfeng and Wang, Shuohang and Wang, Lijuan and Zhu, Chenguang and Zhang, Pengchuan and Yuan, Lu and Peng, Nanyun and Liu, Zicheng and Zeng, Michael},
  booktitle = {The Conference on Computer Vision and Pattern Recognition (CVPR-22)},
  year = {2022}
}

Related Publications

  • Generalized Decoding for Pixel, Image and Language

    Xueyan Zou*, Zi-Yi Dou*, Jianwei Yang*, Zhe Gan, Linjie Li, Chunyuan Li, Xiyang Dai, Harkirat Behl, Jianfeng Wang, Lu Yuan, Nanyun Peng, Lijuan Wang, Yong Jae Lee, and Jianfeng Gao, in The Conference on Computer Vision and Pattern Recognition (CVPR-23), 2023.
    Full Text Code BibTeX Details
    @inproceedings{xdecoder,
      title = {Generalized Decoding for Pixel, Image and Language},
      author = {Zou*, Xueyan and Dou*, Zi-Yi and Yang*, Jianwei and Gan, Zhe and Li, Linjie and Li, Chunyuan and Dai, Xiyang and Behl, Harkirat and Wang, Jianfeng and Yuan, Lu and Peng, Nanyun and Wang, Lijuan and Lee, Yong Jae and Gao, Jianfeng},
      booktitle = {The Conference on Computer Vision and Pattern Recognition (CVPR-23)},
      year = {2023}
    }
    
    Details
  • Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone

    Zi-Yi Dou, Aishwarya Kamath, Zhe Gan, Pengchuan Zhang, Jianfeng Wang, Linjie Li, Zicheng Liu, Ce Liu, Yann LeCun, Nanyun Peng, Jianfeng Gao, and Lijuan Wang, in Proceedings of the Thirty-Sixth Conference on Neural Information Processing Systems (NeurIPS), 2022.
    Full Text BibTeX Details
    @inproceedings{dou2022fiber,
      title = {Coarse-to-Fine Vision-Language Pre-training with Fusion in the Backbone},
      author = {Dou, Zi-Yi and Kamath, Aishwarya and Gan, Zhe and Zhang, Pengchuan and Wang, Jianfeng and Li, Linjie and Liu, Zicheng and Liu, Ce and LeCun, Yann and Peng, Nanyun and Gao, Jianfeng and Wang, Lijuan},
      booktitle = {Proceedings of the Thirty-Sixth Conference on Neural Information Processing Systems (NeurIPS)},
      year = {2022}
    }
    
    Details
  • An Empirical Study of Training End-to-End Vision-and-Language Transformers

    Zi-Yi Dou, Yichong Xu, Zhe Gan, Jianfeng Wang, Shuohang Wang, Lijuan Wang, Chenguang Zhu, Pengchuan Zhang, Lu Yuan, Nanyun Peng, Zicheng Liu, and Michael Zeng, in The Conference on Computer Vision and Pattern Recognition (CVPR-22), 2022.
    Full Text Code Abstract BibTeX Details
    Vision-and-language (VL) pre-training has proven to be highly effective on various VL downstream tasks. While recent work has shown that fully transformer-based VL models can be more efficient than previous region-feature-based methods, their performance on downstream tasks often degrades significantly. In this paper, we present METER, a Multimodal End-to-end TransformER framework, through which we investigate how to design and pre-train a fully transformer-based VL model in an end-to-end manner. Specifically, we dissect the model designs along multiple dimensions: vision encoders (e.g., CLIP-ViT, Swin transformer), text encoders (e.g., RoBERTa, DeBERTa), multimodal fusion module (e.g., merged attention vs. co-attention), architectural design (e.g., encoder-only vs. encoder-decoder), and pre-training objectives (e.g., masked image modeling). We conduct comprehensive experiments and provide insights on how to train a performant VL transformer while maintaining fast inference speed. Notably, our best model achieves an accuracy of 77.64% on the VQAv2 test-std set using only 4M images for pre-training, surpassing the state-of-the-art region-feature-based model by 1.04%, and outperforming the previous best fully transformer-based model by 1.6%.
    @inproceedings{dou2022meter,
      title = {An Empirical Study of Training End-to-End Vision-and-Language Transformers},
      author = {Dou, Zi-Yi and Xu, Yichong and Gan, Zhe and Wang, Jianfeng and Wang, Shuohang and Wang, Lijuan and Zhu, Chenguang and Zhang, Pengchuan and Yuan, Lu and Peng, Nanyun and Liu, Zicheng and Zeng, Michael},
      booktitle = {The Conference on Computer Vision and Pattern Recognition (CVPR-22)},
      year = {2022}
    }
    
    Details