English

SLIP: Self-supervision meets Language-Image Pre-training

Computer Vision and Pattern Recognition 2021-12-24 v1

Abstract

Recent work has shown that self-supervised pre-training leads to improvements over supervised learning on challenging visual recognition tasks. CLIP, an exciting new approach to learning with language supervision, demonstrates promising performance on a wide variety of benchmarks. In this work, we explore whether self-supervised learning can aid in the use of language supervision for visual representation learning. We introduce SLIP, a multi-task learning framework for combining self-supervised learning and CLIP pre-training. After pre-training with Vision Transformers, we thoroughly evaluate representation quality and compare performance to both CLIP and self-supervised learning under three distinct settings: zero-shot transfer, linear classification, and end-to-end finetuning. Across ImageNet and a battery of additional datasets, we find that SLIP improves accuracy by a large margin. We validate our results further with experiments on different model sizes, training schedules, and pre-training datasets. Our findings show that SLIP enjoys the best of both worlds: better performance than self-supervision (+8.1% linear accuracy) and language supervision (+5.2% zero-shot accuracy).

Keywords

Cite

@article{arxiv.2112.12750,
  title  = {SLIP: Self-supervision meets Language-Image Pre-training},
  author = {Norman Mu and Alexander Kirillov and David Wagner and Saining Xie},
  journal= {arXiv preprint arXiv:2112.12750},
  year   = {2021}
}

Comments

Code: https://github.com/facebookresearch/SLIP

R2 v1 2026-06-24T08:30:09.471Z