Robustness Problem

The robustness problem in machine learning focuses on developing models that are resilient to various forms of input perturbations and distribution shifts, aiming to ensure reliable and trustworthy predictions. Current research investigates diverse approaches, including adversarial training, test-time adaptation, and the use of regularization techniques (e.g., gradient regularization, frequency regularization) applied to various architectures such as neural networks (including transformers and sparse networks) and diffusion models. Addressing robustness is crucial for deploying machine learning models safely and effectively in real-world applications, particularly in safety-critical domains like autonomous vehicles and healthcare, where unexpected inputs or data variations can have significant consequences.

Papers

October 25, 2024

Analyzing Neural Network Robustness Using Graph Curvature
Shuhang Tan, Jayson Sia, Paul Bogdan, Radoslav Ivanov
Neural Network Robustness Representation Bottleneck Robustness Problem Ricci Flow Graph Curvature Training Bottleneck

July 5, 2024

Regulating Model Reliance on Non-Robust Features by Smoothing Input Marginal Density
Peiyu Yang, Naveed Akhtar, Mubarak Shah, Ajmal Mian
Feature Attribution Robustness Problem Model Trustworthiness Non Robust Feature

May 31, 2024

Weak Robust Compatibility Between Learning Algorithms and Counterfactual Explanation Generation Algorithms
Ao Xu, Tieru Wu
Native Robustness Counterfactual Explanation Learning Algorithm Probably Approximately Correct Robustness Problem Robust Counterfactuals

March 13, 2024

Adaptive Splitting of Reusable Temporal Monitors for Rare Traffic Violations
Craig Innes, Subramanian Ramamoorthy
Autonomous Vehicle Signal Temporal Logic Splitting Algorithm Robustness Problem Incomplete Trajectory Failure Probability

October 21, 2023

Training Image Derivatives: Increased Accuracy and Universal Robustness
Vsevolod I. Avrutskiy
Robustness Analysis General Robustness Robustness Problem Image Derivative

July 15, 2023

Why Does Little Robustness Help? Understanding and Improving Adversarial Transferability from Surrogate Training
Yechao Zhang, Shengshan Hu, Leo Yu Zhang, Junyu Shi, Minghui Li, Xiaogeng Liu, Wei Wan, Hai Jin
Adversarial Example Adversarial Training Adversarial Sample Robustness Problem Surrogate Model Training

June 16, 2023

Evaluating the Robustness of Text-to-image Diffusion Models against Real-world Attacks
Hongcheng Gao, Hao Zhang, Yinpeng Dong, Zhijie Deng
Native Robustness Text to Image Diffusion Model Text to Image Text Encoder Real World Attack Robustness Problem

June 6, 2023

On Pitfalls of Test-Time Adaptation
Hao Zhao, Yuejiang Liu, Alexandre Alahi, Tao Lin
Distribution Shift Test Time Adaptation Common Pitfall State of the Art Algorithm Robustness Problem

March 11, 2023

Use Perturbations when Learning from Explanations
Juyeon Heo, Vihari Piratla, Matthew Wicker, Adrian Weller
Native Robustness Line by Line Explanation Human Explanation Robustness Problem Smoothing Algorithm Real World Perturbation

March 1, 2023

Combating Exacerbated Heterogeneity for Robust Models in Federated Learning
Jianing Zhu, Jiangchao Yao, Tongliang Liu, Quanming Yao, Jianliang Xu, Bo Han
Adversarial Training Robust Model Adversarial Data Robustness Problem Federated Adversarial

December 24, 2022

Frequency Regularization for Improving Adversarial Robustness
Binxiao Huang, Chaofan Tao, Rui Lin, Ngai Wong
Adversarial Training Adversarial Robustness Adversarial Input Imperceptible Adversarial Perturbation Robustness Problem Frequency Regularization

July 9, 2022

On the Robustness and Anomaly Detection of Sparse Neural Networks
Morgane Ayle, Bertrand Charpentier, John Rachwan, Daniel Zügner, Simon Geisler, Stephan Günnemann
Native Robustness Anomaly Detection Structured Sparsity Sparse Neural Network Sparse Subnetworks Robustness Problem

June 16, 2022

Catastrophic overfitting can be induced with discriminative non-robust features
Guillermo Ortiz-Jiménez, Pau de Jorge, Amartya Sanyal, Adel Bibi, Puneet K. Dokania, Pascal Frossard, Gregory Rogéz, Philip H. S. Torr
Native Robustness Adversarial Training Robust Neural Network Catastrophic Overfitting Robustness Problem Non Robust Feature

February 12, 2022

EREBA: Black-box Energy Testing of Adaptive Neural Networks
Mirazul Haque, Yaswanth Yadlapalli, Wei Yang, Cong Liu
Energy Efficient Robustness Problem Box Testing Adaptive Neural

February 8, 2022

Are Transformers More Robust? Towards Exact Robustness Verification for Transformers
Brian Hsuan-Cheng Liao, Chih-Hong Cheng, Hasan Esen, Alois Knoll
Native Robustness Transformer Megatron Decepticons Robustness Verification Robustness Problem

January 30, 2022

DearFSAC: An Approach to Optimizing Unreliable Federated Learning via Deep Reinforcement Learning
Chenghao Huang, Weilong Chen, Yuxi Chen, Shunji Yang, Yanru Zhang
Deep Reinforcement Learning Native Robustness Constructive Approach Soft Actor Critic Defect Sample Robustness Problem