Research map / Adversarial robustness and security / Backdoor and poisoning
Poisoning-Based Backdoor Attacks in Computer Vision
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add a transformation-based plug-in module to address differences between training and test triggers | Physical-world image-classification backdoor attacks · Image classification, physical world | attack effectiveness | Helped |
| Use invisible, sample-specific triggers instead of a shared trigger | Image-classification backdoor attacks · Image classification, digital world | defense bypass | Helped |
| Use attributes as sample-specific triggers under the clean-label setting | BAAT · Image classification, clean-label setting | attack effectiveness | Mixed |
| Optimize feature-space loss and standard tracking loss alternately | Few-shot backdoor attack (FSBA) · Visual object tracking, digital and physical spaces | attack effectiveness | Helped |
| Use external patterns as backdoor watermarks for ownership verification | Dataset copyright protection · Third-party model ownership verification | ownership verification | Helped |
| Reduce trigger generalization when designing backdoor-based XAI evaluation | Faithful XAI evaluation · Backdoor-based evaluation of XAI methods | evaluation faithfulness | Helped |
Closest papers by meaning
- A Theoretical Analysis of Backdoor Poisoning Attacks in Convolutional Neural Networks (ICML 2024 · similarity 0.91)
- Demystifying Poisoning Backdoor Attacks from a Statistical Perspective (ICLR 2024 · similarity 0.89)
- Adversarial-Inspired Backdoor Defense via Bridging Backdoor and Adversarial Attacks (AAAI 2025 · similarity 0.89)
- Universal Backdoor Attacks (ICLR 2024 · similarity 0.88)
- Temporal-Distributed Backdoor Attack against Video Based Action Recognition (AAAI 2024 · similarity 0.87)
- Efficient Backdoor Attacks for Deep Neural Networks in Real-world Scenarios (ICLR 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.