← Back to Computer Vision
cs.CV

Running clinical motion capture on your smartphone, privately

Lauhitya Reddy, Trisha M. Kesar, Hyeokhyen Kwon

May 20, 2026

Clinical gait analysis normally requires expensive motion-capture labs and trained technicians. AIGaitor runs the entire pipeline—pose estimation, pose optimization, and AI classification—locally on a smartphone using its neural accelerator, with no cloud uploads. On an iPhone 14, it processes 10-second 4K video in 77 seconds, faster than uploading to a high-end cloud GPU when network latency is included. A survey of 74 rehabilitation clinicians found 92% would adopt an accurate, affordable, easy-to-use tool; privacy and operating cost were their top barriers.
Published as AIGaitor: Privacy-preserving and cloud-free motion analysis for everyone, using edge computing arXiv:2605.21421
Read the original paper →