Xingjun Ma

Affiliations

Fudan University

papers

2,359

total citations

papers (22)

$\alpha$-IoU: A Family of Power Intersection over Union Losses for Bounding Box Regression

NEURIPS 2021arXiv

306

citations

Adversarial Camouflage: Hiding Physical-World Attacks With Natural Styles

CVPR 2020arXiv

257

citations

Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make Student Better

ICCV 2021arXiv

130

citations

Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks

NEURIPS 2021arXiv

113

citations

Short-Term and Long-Term Context Aggregation Network for Video Inpainting

ECCV 2020arXiv

citations

CalFAT: Calibrated Federated Adversarial Training with Label Skewness

NEURIPS 2022arXiv

citations

Unlearnable Clusters: Towards Label-Agnostic Unlearnable Examples

CVPR 2023arXiv

citations

Adversarial Prompt Tuning for Vision-Language Models

ECCV 2024arXiv

citations

BlueSuffix: Reinforced Blue Teaming for Vision-Language Models Against Jailbreak Attacks

ICLR 2025arXiv

citations

TAPT: Test-Time Adversarial Prompt Tuning for Robust Inference in Vision-Language Models

CVPR 2025arXiv

citations

Anyattack: Towards Large-scale Self-supervised Adversarial Attacks on Vision-language Models

CVPR 2025arXiv

citations

LDReg: Local Dimensionality Regularized Self-Supervised Learning

ICLR 2024arXiv

citations

IDEATOR: Jailbreaking and Benchmarking Large Vision-Language Models Using Themselves

ICCV 2025arXiv

citations

Free-Form Motion Control: Controlling the 6D Poses of Camera and Objects in Video Generation

ICCV 2025arXiv

citations

AIM: Additional Image Guided Generation of Transferable Adversarial Attacks

AAAI 2025arXiv

citations

HoneypotNet: Backdoor Attacks Against Model Extraction

AAAI 2025arXiv

citations

Towards Million-Scale Adversarial Robustness Evaluation With Stronger Individual Attacks

CVPR 2025arXiv

citations

StolenLoRA: Exploring LoRA Extraction Attacks via Synthetic Data

ICCV 2025arXiv

citations

Gradient Driven Rewards to Guarantee Fairness in Collaborative Machine Learning

NEURIPS 2021

citations

Xingjun Ma

Affiliations

papers (22)

Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks

Anti-Backdoor Learning: Training Clean Models on Poisoned Data

Clean-Label Backdoor Attacks on Video Recognition Models

$\alpha$-IoU: A Family of Power Intersection over Union Losses for Bounding Box Regression

Adversarial Camouflage: Hiding Physical-World Attacks With Natural Styles

Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make Student Better

Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks

Short-Term and Long-Term Context Aggregation Network for Video Inpainting

CalFAT: Calibrated Federated Adversarial Training with Label Skewness

Unlearnable Clusters: Towards Label-Agnostic Unlearnable Examples

Adversarial Prompt Tuning for Vision-Language Models

BlueSuffix: Reinforced Blue Teaming for Vision-Language Models Against Jailbreak Attacks

TAPT: Test-Time Adversarial Prompt Tuning for Robust Inference in Vision-Language Models

Anyattack: Towards Large-scale Self-supervised Adversarial Attacks on Vision-language Models

LDReg: Local Dimensionality Regularized Self-Supervised Learning

IDEATOR: Jailbreaking and Benchmarking Large Vision-Language Models Using Themselves

Free-Form Motion Control: Controlling the 6D Poses of Camera and Objects in Video Generation

AIM: Additional Image Guided Generation of Transferable Adversarial Attacks

HoneypotNet: Backdoor Attacks Against Model Extraction

Towards Million-Scale Adversarial Robustness Evaluation With Stronger Individual Attacks

StolenLoRA: Exploring LoRA Extraction Attacks via Synthetic Data

Gradient Driven Rewards to Guarantee Fairness in Collaborative Machine Learning

papers (22)

Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks

Anti-Backdoor Learning: Training Clean Models on Poisoned Data

Clean-Label Backdoor Attacks on Video Recognition Models

$\alpha$-IoU: A Family of Power Intersection over Union Losses for Bounding Box Regression

Adversarial Camouflage: Hiding Physical-World Attacks With Natural Styles

Revisiting Adversarial Robustness Distillation: Robust Soft Labels Make Student Better

Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks

Short-Term and Long-Term Context Aggregation Network for Video Inpainting

CalFAT: Calibrated Federated Adversarial Training with Label Skewness

Unlearnable Clusters: Towards Label-Agnostic Unlearnable Examples

Adversarial Prompt Tuning for Vision-Language Models

BlueSuffix: Reinforced Blue Teaming for Vision-Language Models Against Jailbreak Attacks

TAPT: Test-Time Adversarial Prompt Tuning for Robust Inference in Vision-Language Models

Anyattack: Towards Large-scale Self-supervised Adversarial Attacks on Vision-language Models

LDReg: Local Dimensionality Regularized Self-Supervised Learning

IDEATOR: Jailbreaking and Benchmarking Large Vision-Language Models Using Themselves

Free-Form Motion Control: Controlling the 6D Poses of Camera and Objects in Video Generation

AIM: Additional Image Guided Generation of Transferable Adversarial Attacks

HoneypotNet: Backdoor Attacks Against Model Extraction

Towards Million-Scale Adversarial Robustness Evaluation With Stronger Individual Attacks

StolenLoRA: Exploring LoRA Extraction Attacks via Synthetic Data

Gradient Driven Rewards to Guarantee Fairness in Collaborative Machine Learning