English

Seeing in Words: Learning to Classify through Language Bottlenecks

Computer Vision and Pattern Recognition 2023-07-04 v1 Artificial Intelligence Computation and Language Machine Learning

Abstract

Neural networks for computer vision extract uninterpretable features despite achieving high accuracy on benchmarks. In contrast, humans can explain their predictions using succinct and intuitive descriptions. To incorporate explainability into neural networks, we train a vision model whose feature representations are text. We show that such a model can effectively classify ImageNet images, and we discuss the challenges we encountered when training it.

Keywords

Cite

@article{arxiv.2307.00028,
  title  = {Seeing in Words: Learning to Classify through Language Bottlenecks},
  author = {Khalid Saifullah and Yuxin Wen and Jonas Geiping and Micah Goldblum and Tom Goldstein},
  journal= {arXiv preprint arXiv:2307.00028},
  year   = {2023}
}

Comments

5 pages, 2 figures, Published as a Tiny Paper at ICLR 2023

R2 v1 2026-06-28T11:19:17.215Z