Dancesport Video Lab
AI-Powered Dance Analysis & Coaching Platform
Professional dance video analysis with real-time pose tracking, coaching feedback, and progress tracking. Built for competitive dancers, coaches, and dance studios who want to improve faster.
Capabilities
Everything under the hood
AI-Powered Pose Tracking
Real-time analysis at 15+ FPS using TensorFlow.js MoveNet. Multi-person tracking, Kalman filtering, and occlusion handling keep every dancer locked in frame — even during complex routines and path crossings.
- Real-time analysis at 15+ FPS using TensorFlow.js MoveNet
- Multi-person tracking for couples and group routines
- Kalman filtering and occlusion handling for fast movement
- >90% ID retention during complex movements and path crossings
Coaching Feedback
Turn raw movement into actionable coaching insights. From posture alignment to balance metrics and footwork precision, every detail is measured and explained in real time.
- Posture analysis: spine alignment, shoulder/hip level, head position
- Balance metrics: center of mass, weight distribution, stability scoring
- Footwork precision: step consistency and placement accuracy
- Movement quality: smoothness, velocity consistency, and jerk detection
- Real-time insights while watching your performance
Session Management
Save every analysis, export professional reports, and track progress over time. Compare sessions side-by-side to spot trends and measure improvement.
- Save sessions for future review
- Export PDF reports with charts and metrics
- Progress tracking over weeks and months
- Comparison tools to monitor trends
Visualization Modes
Choose the right view for every moment — from a simple skeleton overlay to competition-ready displays and advanced analysis mode.
- Simple mode: basic skeleton overlay
- Coach mode: detailed feedback annotations
- Performance mode: competition-ready display
- Analysis mode: advanced metrics and quality indicators
Gallery
See it in action
Under the hood
Technical details
A quick look at the technologies that power Dancesport Video Lab.
Under the hood
Technical details
A quick look at the technologies that power Dancesport Video Lab.
Frontend
- Angular 20 (standalone components)
- TensorFlow.js 4.22 (MoveNet MultiPose Lightning)
- PrimeNG v20 UI components
- RxJS + Socket.IO Client 4.8
- TypeScript
Backend
- Node.js + Express 4.18
- MongoDB data storage
- Socket.IO 4.8 real-time notifications
- Multer file uploads
- JWT authentication
AI / ML
- TensorFlow.js 4.22
- MoveNet MultiPose model
- WebGL / WebGPU backends
- Kalman filtering
- Occlusion recovery algorithms
Use cases
Who it is for
Competitive Dancers
- Analyze technique between private lessons
- Track progress over weeks and months
- Identify weaknesses to focus your training
- Perfect routines with detailed feedback
Dance Coaches
- Scale your coaching to more students
- Remote analysis from anywhere
- Professional reports to share
- Objective metrics to supplement your eye
Dance Studios
- Differentiate with cutting-edge technology
- Attract students with AI-powered coaching
- Add value to classes and packages
- Track student progress systematically
Ready to put Dancesport Video Lab to work?
Tell us what you need and we'll show you how it fits your workflow.
Request a demo