We build tools that give every dancer access to the injury prevention science that was never meant for them.
What is “AXISPHERE”?
AXISPHERE is an early-stage research platform investigating whether markerless computer vision can reliably measure ballet-specific movement using accessible video. Our current work focuses on validating 2D and 3D kinematic estimates, identifying failure modes, and developing transparent, dancer-friendly movement visualizations. AXISPHERE is not currently a diagnostic or injury-prediction tool.
Step 1: Move
Dancers perform movements naturally in a structured session, whether that be from barre exercises to a selected variation.
Step 2: Analyze
AXISPHERE processes movement data in real time using pose estimation, whether that be in person or through recorded videos.
Step 3: Insights
Dancers receive personalized injury risk feedback.
Current Phase
We have entered Phase 2 of AXISPHERE’s development, focused on training and evaluating our 2D pose-estimation model using curated, publicly available movement datasets. Moving away from in-person dancer data collection allows us to build on broader and more standardized sources while continuing to refine the model’s ability to analyze movement patterns relevant to dancer wellness and injury prevention. We are also actively seeking additional funding to advance this work and explore AXISPHERE’s expansion from 2D pose estimation into a fully 3D movement-analysis system.