Paper accepted at ACM CCS 2026: “Illusion of Depth”
Stereo cameras estimate depth from the disparity between corresponding points in the left and right images, underpinning distance perception in autonomous systems and robots that do not carry LiDAR. This work reveals vulnerabilities hidden in that correspondence process.
By projecting patterns onto the scene, an attacker can cause the left and right images to be matched incorrectly and steer the depth estimate in an intended direction. The work will be presented at ACM CCS 2026 in The Hague in November 2026. It is a collaboration among Keio University, The University of Electro-Communications, and the University of Florida.
[Paper]
Sri Hrushikesh Varma Bhupathiraju, Tetsu Ishizue, Nicola Costagliola, Ozora Sako, Kentaro Yoshioka, Takeshi Sugawara, Sara Rampazzi, “Illusion of Depth: Revealing Hidden Stereo Vision Vulnerabilities in Depth Estimation,” Proceedings of the ACM SIGSAC Conference on Computer and Communications Security (ACM CCS 2026), The Hague, Netherlands, November 2026.