Full Professor, Università degli Studi di Genova
2 papers at NeurIPS 2025
By incorporating differentiable constraint evaluation functions directly into stable diffusion’s iterative sampling, this approach enforces physical and functional requirements in generative tasks.
We introduce TransferBench, a comprehensive benchmark for evaluating ensemble-based black-box adversarial attacks under realistic scenarios, revealing limitations in surrogate model choices, robustness generalization, and query efficiency.
