EXIF as Language: Learning Cross-Modal Associations Between Images and Camera Metadata

Chenhao Zheng
Ayush Shrivastava
Andrew Owens

CVPR 2023 (highlight)

[Paper]       [Code]

Quick Jump:

  1. Overview
  2. Cross-modal Training
  3. Applications: Zero-Shot Splice Detection
  4. Qualitative Splice Localization Results


In this paper, we learn visual representations that convey camera properties by creating a joint embedding between image patches and photo metadata. This model treats metadata as a language-like modality: it converts EXIF tags that compose the metadata into a long piece of text, and processes it using an off-the-shelf model from natural language processing. We demonstrate the effectiveness of our learned features on a variety of downstream tasks that require an understanding of low-level imaging properties, where it outperforms other feature representations. In particular, we successfully apply our model to the problem of detecting image splices "zero shot", by clustering the crossmodal embeddings within an image.

Cross-modal Training

Photo metadata for each image (EXIF file) is a freely available but often ignored signal that can be used to learn camera property from images. We propose to treat EXIF tags as language and use multimodal contrative learning techniques to associate images with their metadata.

      Common EXIF tags and example values

Cross-modal image and photo metadata training

Application: Zero-Shot Splice Detection

We apply our learned representation to tasks that require understanding camera properties. For example, we can detect image splicing “zero shot” (and without metadata at test time) by finding inconsistent embeddings within an image. We show a manipulated image that contains content from two source photos. Since these photos were captured with different cameras, the two regions have dissimilar embeddings (visualized by PCA). We localize the splice by clustering the image’s patch embeddings.

Qualitative Splice Localization Results


This research was developed with funding from the Defense Advanced Research Projects Agency (DARPA) under Contract No. HR001120C0123. The views, opinions and/or findings expressed are those of the author and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government. We thank Alexei Efro, Ziyang Chen, Yuexi Du for the helpful discussions. The webpage template was adopted from Colorization project and HOGgles project.

Creative Commons License
This paper is licensed under a Creative Commons Attribution 4.0 International License