Adversarial Detection: Attacking Object Detection in Real Time

September 05, 2022 Β· Declared Dead Β· πŸ› 2023 IEEE Intelligent Vehicles Symposium (IV)

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Han Wu, Syed Yunas, Sareh Rowlands, Wenjie Ruan, Johan Wahlstrom arXiv ID 2209.01962 Category cs.AI: Artificial Intelligence Cross-listed cs.CV, cs.RO Citations 5 Venue 2023 IEEE Intelligent Vehicles Symposium (IV) Last Checked 4 months ago
Abstract
Intelligent robots rely on object detection models to perceive the environment. Following advances in deep learning security it has been revealed that object detection models are vulnerable to adversarial attacks. However, prior research primarily focuses on attacking static images or offline videos. Therefore, it is still unclear if such attacks could jeopardize real-world robotic applications in dynamic environments. This paper bridges this gap by presenting the first real-time online attack against object detection models. We devise three attacks that fabricate bounding boxes for nonexistent objects at desired locations. The attacks achieve a success rate of about 90% within about 20 iterations. The demo video is available at https://youtu.be/zJZ1aNlXsMU.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Artificial Intelligence

Died the same way β€” πŸ‘» Ghosted