English

Pre-training image-language transformers for open-vocabulary tasks

Computer Vision and Pattern Recognition 2022-09-12 v1

Abstract

We present a pre-training approach for vision and language transformer models, which is based on a mixture of diverse tasks. We explore both the use of image-text captioning data in pre-training, which does not need additional supervision, as well as object-aware strategies to pre-train the model. We evaluate the method on a number of textgenerative vision+language tasks, such as Visual Question Answering, visual entailment and captioning, and demonstrate large gains over standard pre-training methods.

Keywords

Cite

@article{arxiv.2209.04372,
  title  = {Pre-training image-language transformers for open-vocabulary tasks},
  author = {AJ Piergiovanni and Weicheng Kuo and Anelia Angelova},
  journal= {arXiv preprint arXiv:2209.04372},
  year   = {2022}
}
R2 v1 2026-06-28T01:01:31.233Z