Structured Light Speckle

Joint egocentric depth estimation and low-latency contact detection via remote vibrometry

ACM UIST 2023
Paul Streli, Jiaxi Jiang, Juliete Rossie, and Christian Holz
Structured Light Speckle teaser image

TapLight is an egocentric remote contact sensing system that simultaneously discovers physical surfaces and moments of touch through a novel integration of structured light and laser speckle: (a) The diffracted laser creates a sparse dot pattern on the surface, producing binocular disparity in the RGB camera’s view, from which we estimate depth values and fit a plane. The monochrome camera’s exposed sensor captures the interference of diffused laser reflections, and its high framerate reveals the remote vibrations propagating through both, the hand and the surface upon contact. TapLight combines detected surfaces with the tracked hand poses from the VR headset, verifies and determines touch locations, and relays them into Virtual Reality (b).

Abstract

Despite advancements in egocentric hand tracking using head-mounted cameras, contact detection with real-world objects remains challenging, particularly for the quick motions often performed during interaction in Mixed Reality. In this paper, we introduce a novel method for detecting touch on discovered physical surfaces purely from an egocentric perspective using optical sensing. We leverage structured laser light to detect real-world surfaces from the disparity of reflections in real-time and, at the same time, extract a time series of remote vibrometry sensations from laser speckle motions. The pattern caused by structured laser light reflections enables us to simultaneously sample the mechanical vibrations that propagate through the user’s hand and the surface upon touch. We integrated Structured Light Speckle into TapLight, a prototype system that is a simple add-on to Mixed Reality headsets. In our evaluation with a Quest 2, TapLight—while moving—reliably detected horizontal and vertical surfaces across a range of surface materials. TapLight also reliably detected rapid touch contact and robustly discarded other hand motions to prevent triggering spurious input events. Despite the remote sensing principle of Structured Light Speckle, our method achieved a latency for event detection in realistic settings that matches body-worn inertial sensing without needing such additional instrumentation. We conclude with a series of VR demonstrations for situated interaction that leverage the quick touch interaction TapLight supports.

Video

Reference

Paul Streli, Jiaxi Jiang, Juliete Rossie, and Christian Holz. Structured Light Speckle: Joint egocentric depth estimation and low-latency contact detection via remote vibrometry. In Proceedings of ACM UIST 2023.

Acknowledgments

We thank NVIDIA for the provision of computing resources through the NVIDIA Academic Grant.