No description

Python 100%

Find a file

tinjin 076efd21ba 更新 ai-motion-feedback.md		2026-05-10 23:32:54 +08:00
ai-motion-feedback.md	更新 ai-motion-feedback.md	2026-05-10 23:32:54 +08:00
ai-pipeline.md	更新 ai-pipeline.md	2026-05-10 23:32:33 +08:00
app.py	添加 app.py	2026-05-10 23:26:52 +08:00
contact-point-detection.md	更新 contact-point-detection.md	2026-05-10 23:29:05 +08:00
full-analysis-workflow.md	更新 full-analysis-workflow.md	2026-05-10 23:32:11 +08:00
metrics.md	更新 metrics.md	2026-05-10 23:31:28 +08:00
motion-recognition.md	更新 motion-recognition.md	2026-05-10 23:30:49 +08:00
README.md	更新 README.md	2026-05-10 23:30:21 +08:00
requirements.txt	更新 requirements.txt	2026-05-10 23:29:49 +08:00
Runnable Prototype	添加 Runnable Prototype	2026-05-09 23:02:17 +08:00
system-architecture.md	更新 system-architecture.md	2026-05-10 23:28:03 +08:00

README.md

RallyMind AI W2

AI Tennis Motion Analysis Prototype

RallyMind AI is a runnable AI prototype for tennis motion analysis.

This project focuses on validating the technical feasibility of AI-powered sports motion intelligence through runnable Skills and a working Prototype.

Core Skills

1. Tennis Motion Recognition

Detects:

Forehand
Backhand
Serve

2. Contact Point Detection

Analyzes:

Swing timing
Contact timing
Body rotation timing

3. AI Motion Feedback

Generates:

Motion correction suggestions
Training advice
Technique feedback

Prototype Workflow

Video Upload → Pose Estimation → Motion Analysis → Contact Point Detection → AI Feedback → Result Output

Technology Stack

MediaPipe
OpenCV
Streamlit
Python
GPT / Gemini API

Runnable Prototype

This repository includes a runnable prototype for AI tennis motion analysis.

Evaluation Metrics

Motion recognition accuracy
Contact point estimation
AI feedback quality
Workflow completion

Future Direction

AI Tennis Coach Agent
Smart Glasses Integration
Motion Intelligence Infrastructure