# Unsupervised Vision-and-Language Pre-training Without Parallel Images and Captions

arXiv: https://arxiv.org/abs/2010.12831 · HTML: https://mlsanity.com/paper/2010.12831/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Oct 2020.

## Built on by (most cited first, 7 of 7)

- [Vision-Language Pre-training: Basics, Recent Advances, and Future Trends](https://mlsanity.com/paper/2210.09263/) (2022)
- [Playing Lottery Tickets with Vision and Language](https://mlsanity.com/paper/2104.11832/) (2021)
- [Detector-Free Weakly Supervised Grounding by Separation](https://mlsanity.com/paper/2104.09829/) (2021)
- [SGEITL: Scene Graph Enhanced Image-Text Learning for Visual Commonsense Reasoning](https://mlsanity.com/paper/2112.08587/) (2021)
- [DiMBERT: Learning Vision-Language Grounded Representations with Disentangled Multimodal-Attention](https://mlsanity.com/paper/2210.16431/) (2022)
- [OVIS: Open-Vocabulary Visual Instance Search via Visual-Semantic Aligned Representation Learning](https://mlsanity.com/paper/2108.03704/) (2021)
- [SA-VQA: Structured Alignment of Visual and Semantic Representations for Visual Question Answering](https://mlsanity.com/paper/2201.10654/) (2022)

## Abstract (arXiv)

Pre-trained contextual vision-and-language (V&L) models have achieved impressive performance on various benchmarks. However, existing models require a large amount of parallel image-caption data for pre-training. Such data are costly to collect and require cumbersome curation. Inspired by unsupervised machine translation, we investigate if a strong V&L representation model can be learned through unsupervised pre-training without image-caption corpora. In particular, we propose to conduct ``mask-and-predict'' pre-training on text-only and image-only corpora and introduce the object tags detected by an object recognition model as anchor points to bridge two modalities. We find that such a simple approach achieves performance close to a model pre-trained with aligned data, on four English V&L benchmarks. Our work challenges the widely held notion that aligned data is necessary for V&L pre-training, while significantly reducing the amount of supervision needed for V&L models.