Abstract
Extended reality (XR) input systems relying solely on optical hand-tracking face critical limitations, including contact ambiguity between hovering and touching, processing pipeline latency, and field-of-view occlusion. This disclosure describes a multimodal sensor fusion framework that addresses these challenges by integrating spatial coordinate mapping from head-worn devices with high-frequency inertial event detection from wrist-worn devices. When a wrist-worn sensor detects the sharp acceleration of a physical impact at a specific timestamp, the system triggers retroactive event gating. The system accesses historical visual buffers to retrieve spatial coordinates corresponding to the impact time, then applies an update that dynamically forces the depth uncertainty to zero. Alternatively, synchronized multimodal streams can feed directly into deep learning models. This mechanism snaps the estimated position onto the contact surface, resolving contact ambiguity and eliminating perceived latency.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Chugh, Tushar; Mone, Aditya Shrikant; and Kumara, Karthik, "Gating Spatial Coordinates Utilizing Inertial Detection", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/10423