yanpeigong

yanpeigong / Evosmash

Public

[National College Student Software Innovation Competition] 🏸AI-powered badminton analysis, coaching, and tactical evolution agent

agent deep-learning software
33
0
69% credibility
Found Apr 20, 2026 at 33 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

EvoSmash is a mobile app that analyzes badminton rally videos using computer vision and AI to deliver tactical coaching advice that evolves with user feedback.

How It Works

1
📱 Open EvoSmash on your phone

You launch the badminton coaching app ready to improve your game.

2
⚙️ Pick your game mode

Choose singles or doubles so it understands your court setup.

3
🎥 Record or upload a rally clip

Capture a quick video of your rally or pick one from your gallery, feeling the excitement build.

4
🔍 Watch the magic analyze

The app scans the shuttle path, your movement, and the court to break down what happened.

5
💡 Get instant coaching tips

See smart tactic suggestions, speed stats, and personalized advice on your next move.

6
📈 Review your progress

Check your library of past rallies and watch your skills evolve over time.

🏆 Become a smarter player

Your game improves with every clip as the coach learns from your wins and losses.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 33 to 33 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is Evosmash?

Evosmash is a Python-based AI agent for badminton video analysis, turning raw rally or match footage into actionable coaching insights. Upload a clip via its FastAPI backend or React web/Android app, and it delivers physics breakdowns like shuttle speed and landing calls, plus tactical recommendations from an evolving LLM coach. It solves the pain of manual game review by automating trajectory tracking, pose analysis, and self-improving tactics through feedback loops, much like a national college innovation project on agent-driven sports analysis.

Why is it gaining traction?

It stands out with a closed-loop agent that adapts tactics using Bayesian updates on your feedback, unlike static analyzers—your wins/losses refine future advice on pressure plays or net control. Developers dig the end-to-end pipeline: video upload yields instant reports with duel simulations and training blocks, deployable as a web app or mobile via Capacitor. The badminton niche hooks sports AI tinkerers seeking real-world CV and agent examples beyond generic github national geographic iptv github tools.

Who should use this?

Badminton players reviewing personal matches for tactical edges, coaches at national college grounds prescribing drills from rally footage, or sports devs prototyping analysis agents. Ideal for club teams tired of spreadsheets, or indie hackers building fitness apps with video AI.

Verdict

Grab it if you're into sports tech—33 stars show early promise from a national college competition, but the 0.7% credibility score flags immaturity like sparse docs and tests. Fork and contribute to mature this agent for broader analysis use cases.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.