Adversarial Alignment

Adversarial alignment is a technique used in machine learning to reduce discrepancies between data distributions from different domains, enabling improved model generalization and robustness. Current research focuses on applying adversarial alignment within various frameworks, including unsupervised domain adaptation, few-shot learning, and generative adversarial networks (GANs), often incorporating active learning or other regularization methods to enhance performance and mitigate negative transfer. This approach is crucial for addressing challenges in diverse applications such as fault diagnosis, medical image reconstruction, and text classification, where data scarcity or domain shifts hinder model accuracy and reliability. The ultimate goal is to develop more robust and adaptable machine learning models capable of handling real-world data complexities.

Papers

November 4, 2024

Alignment-Based Adversarial Training (ABAT) for Improving the Robustness and Accuracy of EEG-Based BCIs
Xiaoqing Chen, Ziwei Wang, Dongrui Wu
Adversarial Attack Native Robustness Brain Computer Interface Motor Imagery Classification Adversarial Alignment

September 27, 2024

A3: Active Adversarial Alignment for Source-Free Domain Adaptation
Chrisantus Eze, Christopher Crick
Adversarial Training Unsupervised Domain Adaptation Adversarial Learning Source Free Domain Adaptation Adversarial Alignment

August 19, 2024

Coarse-Fine View Attention Alignment-Based GAN for CT Reconstruction from Biplanar X-Rays
Zhi Qiao, Hanqiang Ouyang, Dongheng Chu, Huishu Yuan, Xiantong Zhen, Pei Dong, Zhen Qian
Feature Fusion X Ray Computed Tomography Reconstruction Adversarial Alignment Coarse Grained Attention

May 25, 2024

Overcoming Negative Transfer by Online Selection: Distant Domain Adaptation for Fault Diagnosis
Ziyan Wang, Mohamed Ragab, Wenmian Yang, Min Wu, Sinno Jialin Pan, Jie Zhang, Zhenghua Chen
Domain Adaptation Unsupervised Domain Adaptation Conditional Generative Adversarial Network Unlabeled Target Domain Negative Transfer Adversarial Alignment

May 6, 2024

Generative adversarial learning with optimal input dimension and its adaptive generator architecture
Zhiyao Tan, Ling Zhou, Huazhen Lin
Generative Adversarial Network Adversarial Learning Generative Question Efficient Generation Adversarial Alignment Input Dimension

December 18, 2023

Regularized Conditional Alignment for Multi-Domain Text Classification
Juntao Hu, Yuan Wu
Adversarial Training Domain Invariant Feature Optimal Alignment Adversarial Alignment Domain Text Classification

September 5, 2023

Dual Adversarial Alignment for Realistic Support-Query Shift Few-shot Learning
Siyang Jiang, Rui Fang, Hsi-Wen Chen, Wei Ding, Ming-Syan Chen
Shot Learning Customer Service Adversarial Alignment Adversarial Feature Alignment

June 26, 2023

Are aligned neural networks adversarially aligned?
Nicholas Carlini, Milad Nasr, Christopher A. Choquette-Choo, Matthew Jagielski, Irena Gao, Anas Awadalla, Pang Wei Koh, Daphne Ippolito, Katherine Lee, Florian Tramer, Ludwig Schmidt
Neural Network Adversarial Example Adversarial Input Adversarial Alignment

June 5, 2023

Adversarial alignment: Breaking the trade-off between the strength of an attack and its relevance to human perception
Drew Linsley, Pinyuan Feng, Thibaut Boissin, Alekh Karkada Ashok, Thomas Fel, Stephanie Olaiya, Thomas Serre
Deep Learning Deep Neural Network Adversarial Attack Adversarial Robustness Adversarial Perturbation New Attack Relative Relevance Human Perception Adversarial Alignment

May 29, 2023

Conditional Support Alignment for Domain Adaptation with Label Shift
Anh T Nguyen, Lam Tran, Anh Tong, Tuan-Duy H. Nguyen, Toan Tran
Domain Adaptation Unsupervised Domain Adaptation Label Shift Adversarial Alignment

January 11, 2023

Adversarial Alignment for Source Free Object Detection
Qiaosong Chu, Shuyan Li, Guangyi Chen, Kai Li, Xiu Li
Adversarial Learning Trained Source Model Cross Domain Object Detection Label Rich Source Adversarial Alignment Source Free Object Detection

September 9, 2022

PoliTO-IIT-CINI Submission to the EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge for Action Recognition
Mirco Planamente, Gabriele Goletto, Gabriele Trivigno, Giuseppe Averta, Barbara Caputo
Action Recognition Multi Source Domain Adaptation UniUD FBK UB UniBZ Submission Adversarial Alignment

May 8, 2022

PGADA: Perturbation-Guided Adversarial Alignment for Few-shot Learning Under the Support-Query Shift
Siyang Jiang, Wei Ding, Hsi-Wen Chen, Ming-Syan Chen
Shot Learning Optimal Transport Customer Service Adversarial Alignment Adversarial Style Augmentation

March 16, 2022

Adversarial Support Alignment
Shangyuan Tong, Timur Garipov, Yang Zhang, Shiyu Chang, Tommi S. Jaakkola
Alignment Problem Label Distribution Distribution Alignment Domain Adaptation Task Adversarial Alignment

March 3, 2022

Robustness and Adaptation to Hidden Factors of Variation
William Paul, Philippe Burlina
Generative Model Native Robustness Strong Generalization Adaptation Concern Category Wise Variation Robust AI Adversarial Alignment

February 9, 2022

Amplitude Spectrum Transformation for Open Compound Domain Adaptive Semantic Segmentation
Jogendra Nath Kundu, Akshay Kulkarni, Suvaansh Bhambri, Varun Jampani, R. Venkatesh Babu
Domain Adversarial Open Set Domain Adaptation Adversarial Alignment Domain Style

Adversarial Alignment

Papers

Alignment-Based Adversarial Training (ABAT) for Improving the Robustness and Accuracy of EEG-Based BCIs

A3: Active Adversarial Alignment for Source-Free Domain Adaptation

Coarse-Fine View Attention Alignment-Based GAN for CT Reconstruction from Biplanar X-Rays

Overcoming Negative Transfer by Online Selection: Distant Domain Adaptation for Fault Diagnosis

Generative adversarial learning with optimal input dimension and its adaptive generator architecture

Regularized Conditional Alignment for Multi-Domain Text Classification

Dual Adversarial Alignment for Realistic Support-Query Shift Few-shot Learning

Are aligned neural networks adversarially aligned?

Adversarial alignment: Breaking the trade-off between the strength of an attack and its relevance to human perception

Conditional Support Alignment for Domain Adaptation with Label Shift

Adversarial Alignment for Source Free Object Detection

PoliTO-IIT-CINI Submission to the EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge for Action Recognition

PGADA: Perturbation-Guided Adversarial Alignment for Few-shot Learning Under the Support-Query Shift

Adversarial Support Alignment

Robustness and Adaptation to Hidden Factors of Variation

Amplitude Spectrum Transformation for Open Compound Domain Adaptive Semantic Segmentation