{"slug": "gaitvista-reliability-aware-ai-measurement-toward-accessible-longitudinal-gait", "title": "GaitVista: Reliability-Aware AI Measurement toward Accessible Longitudinal Gait Assessment", "summary": "Researchers introduced GaitVista, a reliability-aware measurement layer that reduces average full-body and lower-body gait measurement error by 27.7% and 27.8% across seven clean and degraded sensing conditions on the TotalCapture benchmark. GaitVista uses a lightweight gate that assigns joint- and frame-specific visual contributions based on camera coverage, local visual quality, cross-modal disagreement, and root-motion continuity, cutting the gap to a joint-frame oracle from 2.76–5.33 cm for condition-blind baselines to 1.11 cm. On MoVi with image-derived keypoints, GaitVista reduced marker-supported error by 6.4% versus the strongest learned fusion baseline and improved bilateral knee-flexion waveform accuracy by 18.9% on TotalCapture, though the authors report progress toward accessible gait assessment rather than validated clinical deployment.", "body_md": "arXiv:2609.22619v1 Announce Type: new \nAbstract: Tracking recovery of walking function requires detecting meaningful gait change across rehabilitation sessions, yet objective 3D measurement remains confined to specialized motion-capture laboratories. Small camera sets and body-worn inertial sensors broaden access, but reliability varies across joints and time, allowing sensing failures to masquerade as patient change. We present \\textsc{GaitVista}, a reliability-aware measurement layer whose lightweight gate assigns joint- and frame-specific visual contributions using camera coverage, local visual quality, cross-modal disagreement, and root-motion continuity, and exposes them for inspection. Across seven clean and degraded sensing conditions on TotalCapture, \\textsc{GaitVista} reduces average full-body and lower-body error by \\textbf{27.7\\%} and \\textbf{27.8\\%}, attains the lowest worst-condition error among fusion methods, and reduces the gap to a joint-frame oracle from $2.76$--$5.33$~cm for condition-blind baselines to $1.11$~cm. On MoVi with image-derived keypoints, it is the only deployable fusion method to improve over both unimodal streams, reducing marker-supported error by \\textbf{6.4\\%} relative to the strongest learned fusion baseline. On TotalCapture, it improves bilateral knee-flexion waveform accuracy by \\textbf{18.9\\%}. Raw inertial measurements from five TotalCapture participants show location- and time-varying magnetic disturbance, supporting the design's reliability premise. Both benchmarks contain neurologically healthy participants in controlled settings and retain participant-specific IMU calibration; we therefore report progress toward accessible gait assessment, not validated clinical deployment.", "url": "https://wpnews.pro/news/gaitvista-reliability-aware-ai-measurement-toward-accessible-longitudinal-gait", "canonical_source": "https://arxiv.org/abs/2609.22619", "published_at": "2026-09-23 04:00:00+00:00", "updated_at": "2026-09-23 04:26:48.753926+00:00", "lang": "en", "topics": ["machine-learning", "computer-vision", "ai-research", "ai-safety"], "entities": ["GaitVista", "TotalCapture", "MoVi", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/gaitvista-reliability-aware-ai-measurement-toward-accessible-longitudinal-gait", "markdown": "https://wpnews.pro/news/gaitvista-reliability-aware-ai-measurement-toward-accessible-longitudinal-gait.md", "text": "https://wpnews.pro/news/gaitvista-reliability-aware-ai-measurement-toward-accessible-longitudinal-gait.txt", "jsonld": "https://wpnews.pro/news/gaitvista-reliability-aware-ai-measurement-toward-accessible-longitudinal-gait.jsonld"}}