Skip to content
Assessing the security posture of a widely used vision model: YOLOv7
blog.trailofbits.com | vulnerability | #rce | #open-source | #denial-of-service | #deserialization | #vulnerability | #machine-learning | #computer-vision | #trail-of-bits | #model-security | #yolov7
Summary
Trail of Bits found 11 vulnerabilities in YOLOv7, including YAML parsing that enables remote code execution, denial of service, and model differentials, deeming it unfit for mission-critical use.
- Published
- Collected