We introduce Angle-Robust Concept Learning (AngleRoCL), a simple and flexible approach that learns a generalizable concept representing the capability of generating angle-robust patches. The learne…

An open source implementation of CLIP.

[AAAI2022] Code Release of Attacking Video Recognition Models with Bullet-Screen Comments

[GRSM] Project Page for "GeoPix: Multi-Modal Large Language Model for Pixel-level Image Understanding in Remote Sensing"

📚 从零开始的智能体原理与实践教程

[ACM MM 2024] ReToMe-VA: Recursive Token Merging for Video Diffusion-based Unrestricted Adversarial Attack

A collection of resources on attacks and defenses targeting text-to-image diffusion models

[BMVC 2023] Semantic Adversarial Attacks via Diffusion Models

A data augmentations library for audio, image, text, and video.

Code for NeurIPS 2019 Paper

DeepGauge: Multi-Granularity Testing Criteria for Deep Learning Systems

This project is a reimplementation of The Sketchy Database: Learning to Retrieve Badly Drawn Bunnies

📚 A collection of sketch based application papers.

PaddleSleeve

[VLM-Attack-Survey-2024] Paper list and projects for VLM attacks

😎 up-to-date & curated list of awesome Attacks on Large-Vision-Language-Models papers, methods & resources.

[NeurIPS-2023] Annual Conference on Neural Information Processing Systems

Code for Semantic-Aligned Adversarial Evolution Triangle for High-Transferability Vision-Language Attack(TPAMI 2025)

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

We introduce a novel approach to counter adversarial attacks, namely, image resampling. The underlying rationale behind our idea is that image resampling can alleviate the influence of adversarial …

[CVPR 2022] Official PyTorch Implementation for DiffusionCLIP: Text-guided Image Manipulation Using Diffusion Models

Revisiting Transferable Adversarial Images (TPAMI 2025)

Code for AAAI 2024 paper: CR-SAM: Curvature Regularized Sharpness-Aware Minimization

OpenMMLab Detection Toolbox and Benchmark

[TGRS🔥] Towards Generic and Controllable Attacks Against Object Detection

[ECCV 2024] Official implementation of the paper "Semantic-SAM: Segment and Recognize Anything at Any Granularity"

Data-free Universal Adversarial Perturbation with Pseudo-semantic Prior