# ZK-IMG: Attested Images via Zero-Knowledge Proofs to Fight Disinformation

**Authors:** Daniel Kang, Tatsunori Hashimoto, Ion Stoica, Yi Sun  
**arXiv:** [2211.04775](https://arxiv.org/abs/2211.04775)  
**Date:** 2022-11-09  
**Summary page:** https://www.summarizepaper.com/en/paper/2211.04775v2/

## Key points

- The rise of deepfakes has raised concerns about misinformation and social disruption.
- Researchers propose using ZK-SNARKs and attested cameras to verify image authenticity.
- Existing methods have limitations like lack of input privacy preservation, slow processing speeds for limited image sizes, and reliance on custom cryptographic arguments.
- A new solution called zk-img has been introduced to address these challenges.
- zk-img is a library that conceals the original image while attesting to transformations using ZK-SNARKs.
- It enables developers to specify high-level image transformations seamlessly compiled into ZK-SNARKs.
- The tool protects input and output image privacy by computing hashes within the ZK-SNARK itself.
- zk-img offers secure methods for chaining multiple transformations together for unlimited alterations.
- It is the first system capable of securely transforming high-definition images on standard hardware.
- This approach not only tackles deepfake challenges but also enhances verification and authentication in image manipulation applications.

## Summary

The rise of deepfakes has sparked concerns about their potential for spreading misinformation and causing social disruption. To combat this issue, researchers have proposed using ZK-SNARKs (zero-knowledge succinct non-interactive argument of knowledge) and attested cameras to verify the authenticity of images captured by a camera. While ZK-SNARKs offer a way to verify image transformations without interaction and based on standard cryptographic assumptions, existing methods have limitations such as lack of input privacy preservation, slow processing speeds limited to 128x128 images, and reliance on custom cryptographic arguments. In response to these challenges, a new solution called zk-img has been introduced. is a library designed to attest to image transformations while concealing the original image. This tool enables application developers to specify high-level image transformations, which are then compiled into ZK-SNARKs seamlessly. To protect the privacy of input and output images,  computes hashes within the ZK-SNARK itself. Additionally,  offers secure and private methods for chaining multiple image transformations together, allowing for an unlimited number of alterations. By implementing these optimizations,  stands out as the first system capable of transforming high-definition (HD) images on standard hardware securely and privately. This innovative approach not only addresses the challenges posed by deepfakes but also paves the way for more reliable verification and authentication processes in image manipulation applications.