Skip to main content
Cornell University
We gratefully acknowledge support from
the Simons Foundation and member institutions.
arxiv logo > cs > arXiv:2103.00020

Help | Advanced Search

Computer Science > Computer Vision and Pattern Recognition

(cs)
[Submitted on 26 Feb 2021]

Title:Learning Transferable Visual Models From Natural Language Supervision

Authors:Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever
Download PDF
Abstract: State-of-the-art computer vision systems are trained to predict a fixed set of predetermined object categories. This restricted form of supervision limits their generality and usability since additional labeled data is needed to specify any other visual concept. Learning directly from raw text about images is a promising alternative which leverages a much broader source of supervision. We demonstrate that the simple pre-training task of predicting which caption goes with which image is an efficient and scalable way to learn SOTA image representations from scratch on a dataset of 400 million (image, text) pairs collected from the internet. After pre-training, natural language is used to reference learned visual concepts (or describe new ones) enabling zero-shot transfer of the model to downstream tasks. We study the performance of this approach by benchmarking on over 30 different existing computer vision datasets, spanning tasks such as OCR, action recognition in videos, geo-localization, and many types of fine-grained object classification. The model transfers non-trivially to most tasks and is often competitive with a fully supervised baseline without the need for any dataset specific training. For instance, we match the accuracy of the original ResNet-50 on ImageNet zero-shot without needing to use any of the 1.28 million training examples it was trained on. We release our code and pre-trained model weights at this https URL.
Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
Cite as: arXiv:2103.00020 [cs.CV]
  (or arXiv:2103.00020v1 [cs.CV] for this version)
  https://doi.org/10.48550/arXiv.2103.00020
arXiv-issued DOI via DataCite

Submission history

From: Jong Wook Kim [view email]
[v1] Fri, 26 Feb 2021 19:04:58 UTC (6,174 KB)
Full-text links:

Download:

  • PDF
  • Other formats
(license)
Current browse context:
cs.CV
< prev   |   next >
new | recent | 2103
Change to browse by:
cs
cs.LG

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar

3 blog links

(what is this?)

DBLP - CS Bibliography

listing | bibtex
Alec Radford
Jong Wook Kim
Aditya Ramesh
Gabriel Goh
Girish Sastry
…
a export bibtex citation Loading...

Bookmark

BibSonomy logo Mendeley logo Reddit logo ScienceWISE logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack