Official repository for our work on micro-budget training of large-scale diffusion models.

COALA: A Practical and Vision-Centric Federated Learning Platform, accepted to ICML'24

Query-Efficient Data-Free Learning from Black-Box Models

This repo is the official implementation of the ICLR'23 paper "Towards Robustness Certification Against Universal Perturbations." We calculate the certified robustness against universal perturbatio…

The first comprehensive Robustness investigation benchmark on large-scale dataset ImageNet regarding ARchitecture design and Training techniques towards diverse noises.