OEM Finder: Hunting Vulnerable OEM IoT Devices at Scale
summary
A tool named OEM Finder analyzes image keypoints with a Cazza-style object recognition, Manhattan-distance matching, relative neighborhood graphs and a shortest-path graph kernel and finds more than 180 unique vulnerable OEM device candidates from over 54,000 network camera images scraped from Amazon and Walmart. The team downloads OEM vendor firmware to confirm matches and reproduces vulnerabilities — e.g., Dahua CVE-2017-9315 hard-coded credentials — in multiple rebranded firmwares. OEM Finder is published as an online search engine.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:04OEM security risks
- 3:31Research approach
- 5:00Technical methodology
- 7:19Experimental results
- 9:12Case studies
- 12:26Vulnerability analysis
- 13:55OEM finder tool demo
- 17:50Conclusion and takeaways
related talks
45mFirmware.RE: Firmware Unpacking, Analysis and Vulnerability-Discovery as a Service
40mSecurity/Privacy in the IoT generation
52mIoT Hacking 101: Reverse Engineering The Xiaomi Ecosystem
37mFinding Vulnerabilities in IoT Embedded Devices Using Linux OS and Open Source Tools
49mEstablishing and Maintaining Device Trust: A Case Study
33mThe Various Shades of Supply Chain: SBOM, N-Days and Zero Trust
23mCameras, CACs & Clocks: A Story of Millions of Interrogated and...
34mPwning Smart Weighing Machines wt API & Hardware Hacking