Hanxun Huang

[ICML2020] Normalized Loss Functions for Deep Learning with Noisy Labels

HanxunH

HanxunH

[NeurIPS2021] Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks

HanxunH

Reproduce Results for ICCV2019 "Symmetric Cross Entropy for Robust Learning with Noisy Labels" https://arxiv.org/abs/1908.06112

HanxunH

[ICLR2021] Unlearnable Examples: Making Personal Data Unexploitable

HanxunH

[Machine Learning 2023] Imbalanced Gradients: A Subtle Cause of Overestimated Adversarial Robustness

HanxunH

[ICLR2023] Distilling Cognitive Backdoor Patterns within an Image

HanxunH