Summary of Breaking the Illusion: Real-world Challenges For Adversarial Patches in Object Detection, by Jakob Shack et al.
Breaking the Illusion: Real-world Challenges for Adversarial Patches in Object Detection
by Jakob Shack, Katarina Petrovic, Olga Saukh
First submitted to arxiv on: 23 Oct 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The study investigates the effectiveness of adversarial patches against YOLO object detection networks in physical scenarios. Two types of patches are tested: a global patch that can be placed anywhere and a local patch designed to target specific objects. The performance is analyzed under various factors such as patch size, position, rotation, brightness, and hue. The results show that these parameters significantly impact the patch’s effectiveness, highlighting the challenges in maintaining attack efficacy in real-world conditions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Adversarial attacks can harm machine learning systems. This study looks at how well patches work against object detection networks in the real world. Two types of patches are tested: one that can be placed anywhere and another that targets specific objects. The researchers see what happens when they change things like patch size, position, and color. They find that these changes make a big difference in how well the patches work. This shows that we need to think about how real-world conditions affect attacks, so we can make better defenses. |
Keywords
» Artificial intelligence » Machine learning » Object detection » Yolo
