English

UniT: Multimodal Multitask Learning with a Unified Transformer

Computer Vision and Pattern Recognition 2021-08-19 v3 Computation and Language

Abstract

We propose UniT, a Unified Transformer model to simultaneously learn the most prominent tasks across different domains, ranging from object detection to natural language understanding and multimodal reasoning. Based on the transformer encoder-decoder architecture, our UniT model encodes each input modality with an encoder and makes predictions on each task with a shared decoder over the encoded input representations, followed by task-specific output heads. The entire model is jointly trained end-to-end with losses from each task. Compared to previous efforts on multi-task learning with transformers, we share the same model parameters across all tasks instead of separately fine-tuning task-specific models and handle a much higher variety of tasks across different domains. In our experiments, we learn 7 tasks jointly over 8 datasets, achieving strong performance on each task with significantly fewer parameters. Our code is available in MMF at https://mmf.sh.

Keywords

Cite

@article{arxiv.2102.10772,
  title  = {UniT: Multimodal Multitask Learning with a Unified Transformer},
  author = {Ronghang Hu and Amanpreet Singh},
  journal= {arXiv preprint arXiv:2102.10772},
  year   = {2021}
}

Comments

16 pages

R2 v1 2026-06-23T23:23:06.130Z