← All publications

BackdoorBench: A Comprehensive Benchmark and Analysis of Backdoor Learning

International Journal of Computer Vision, 2025

Abstract

BackdoorBench provides a unified, extensible platform for implementing and evaluating backdoor attacks and defenses. This extended study substantially expands the original NeurIPS benchmark with more algorithms, evaluation pairs, analytical perspectives, and practical tools for reproducible research.

Citation

Wu, Baoyuan, et al. "BackdoorBench: A Comprehensive Benchmark and Analysis of Backdoor Learning." International Journal of Computer Vision 133, 5700–5787, 2025.