NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild

2026-04-13

AI Summary

This work addresses the significant performance degradation of existing AI-generated image detectors under real-world transformations such as cropping, resizing, and compression. To advance robust detection research, the authors construct a large-scale benchmark dataset comprising 185,750 AI-generated images from 42 generative models and 108,750 real images, systematically subjected to 36 realistic perturbations. They establish a standardized evaluation protocol using ROC AUC to assess detector performance on both pristine and transformed images. The benchmark attracted 511 participants from 20 teams who submitted valid solutions, substantially advancing the understanding and development of robust detection methods under complex real-world conditions.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Large-scale security measurements

Abstract

This paper presents an overview of the NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild, held in conjunction with the NTIRE workshop at CVPR 2026. The goal of this challenge was to develop detection models capable of distinguishing real images from generated ones in realistic scenarios: the images are often transformed (cropped, resized, compressed, blurred) for practical usage, and therefore, the detection models should be robust to such transformations. The challenge is based on a novel dataset consisting of 108,750 real and 185,750 AI-generated images from 42 generators comprising a large variety of open-source and closed-source models of various architectures, augmented with 36 image transformations. Methods were evaluated using ROC AUC on the full test set, including both transformed and untransformed images. A total of 511 participants registered, with 20 teams submitting valid final solutions. This report provides a comprehensive overview of the challenge, describes the proposed solutions, and can be used as a valuable reference for researchers and practitioners in increasing the robustness of the detection models to real-world transformations.

Problem

• AI-generated image detection
• robustness
• image transformations
• real-world scenarios
• deepfake detection

Innovation

• robust detection
• AI-generated image
• image transformation
• large-scale dataset
• cross-generator generalization