Intrinsic Robustness

Intrinsic robustness in machine learning focuses on developing systems that maintain performance despite inherent uncertainties or variations in input data, rather than relying solely on external defenses. Current research investigates this through various approaches, including adversarial training, improved data augmentation techniques, and analyses of model architectures like neural ordinary differential equations and transformers, aiming to understand and enhance the inherent resilience of models. This research is crucial for building reliable and dependable AI systems across diverse applications, from robotics and natural language processing to computer vision and remote sensing, where real-world data is inherently noisy and unpredictable.

Papers

November 15, 2024

Toward Robust and Accurate Adversarial Camouflage Generation against Vehicle Detectors
Jiawei Zhou, Linye Lyu, Daojing He, Yu Li
Adversarial CAmouflage Intrinsic Robustness Adversarial Texture Camouflage Generation

August 19, 2024

A Biologically Inspired Design Principle for Building Robust Robotic Systems
Xing Li, Oussama Zenkri, Adrian Pfisterer, Oliver Brock
Native Robustness Robotic System Long Horizon Manipulation Task Intrinsic Robustness Inspired Design Robust Robot Adaptive Robotics

March 12, 2024

Toward An Analytic Theory of Intrinsic Robustness for Dexterous Grasping
Albert H. Li, Preston Culbertson, Aaron D. Ames
Dexterous Grasping Closure Model Grasp Planning Intrinsic Robustness Robust Grasp Uncertain Obstacle

February 9, 2024

Fight Back Against Jailbreaking via Prompt Adversarial Tuning
Yichuan Mo, Yuji Wang, Zeming Wei, Yisen Wang
Jailbreak Attack Intrinsic Robustness

January 26, 2024

Better Representations via Adversarial Training in Pre-Training: A Theoretical Perspective
Yue Xing, Xiaofeng Lin, Qifan Song, Yi Xu, Belinda Zeng, Guang Cheng
Adversarial Training Adversarial Robustness Pre Trained Model Pre Training Better Representation Intrinsic Robustness Large Scale Deep Learning

October 31, 2023

Dynamic Batch Norm Statistics Update for Natural Robustness
Shahbaz Rezaei, Mohammad Sadegh Norouzzadeh
Data Augmentation Corruption Level Corruption Robustness Intrinsic Robustness Imagenet C Batchnorm Minus Implementation Batch Norm

August 4, 2023

RobustMQ: Benchmarking Robustness of Quantized Models
Yisong Xiao, Aishan Liu, Tianyuan Zhang, Haotong Qin, Jinyang Guo, Xianglong Liu
Native Robustness Adversarial Robustness Quantization Model Intrinsic Robustness Quantization Robust Parameter

June 21, 2023

A Comprehensive Study on the Robustness of Image Classification and Object Detection in Remote Sensing: Surveying and Benchmarking
Shaohui Mei, Jiawei Lian, Xiaofei Wang, Yuru Su, Mingyang Ma, Lap-Pui Chau
Deep Neural Network Native Robustness Adversarial Robustness Image Classification Remote Sensing Benchmark Platform Adversarial Noise Intrinsic Robustness

May 24, 2023

Non-adversarial Robustness of Deep Learning Methods for Computer Vision
Gorana Gojić, Vladimir Vincan, Ognjen Kundačina, Dragiša Mišković, Dinu Dragan
Deep Learning Native Robustness Computer Vision Adversarial Robustness Better Robustness Intrinsic Robustness

May 18, 2023

Quantifying the robustness of deep multispectral segmentation models against natural perturbations and data poisoning
Elise Bishoff, Charles Godfrey, Myles McKay, Eleanor Byler
Adversarial Attack Native Robustness Adversarial Robustness Additive Perturbation Data Poisoning Multispectral Model Intrinsic Robustness Natural Robustness Research

May 16, 2023

Ortho-ODE: Enhancing Robustness and of Neural ODEs against Adversarial Attacks
Vishal Purohit
Adversarial Attack Native Robustness Deep Learning Model Neural ODE Neural Ordinary Differential Equation Intrinsic Robustness High Order Solver

February 28, 2023

A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking
Chang Liu, Yinpeng Dong, Wenzhao Xiang, Xiao Yang, Hang Su, Jun Zhu, Yuefeng Chen, Yuan He, Hui Xue, Shibao Zheng
Adversarial Attack Native Robustness Adversarial Robustness Benchmark Platform Comprehensive Study Image Classification Model Intrinsic Robustness Robustness Evaluation Benchmark

February 5, 2023

Leaving Reality to Imagination: Robust Classification via Generated Datasets
Hritik Bansal, Aditya Grover
Generative Model Real World Data Generation Robust Classification Intrinsic Robustness Standard Convolutional ImageNet Classifier

June 16, 2022

Noisy Learning for Neural ODEs Acts as a Robustness Locus Widening
Martin Gonzalez, Hatem Hajri, Loic Cantat, Mihaly Petreczky
Native Robustness Neural ODE Enhanced Robustness Neural CDEs Intrinsic Robustness

March 16, 2022

COPA: Certifying Robust Policies for Offline Reinforcement Learning against Poisoning Attacks
Fan Wu, Linyi Li, Chejian Xu, Huan Zhang, Bhavya Kailkhura, Krishnaram Kenthapadi, Ding Zhao, Bo Li
Reinforcement Learning Offline Reinforcement Learning Poisoning Attack Robust Policy Action Repetition Intrinsic Robustness Decentralized Aggregation