CricVision AI
Link to open source: https://github.com/rajat190/cricvision-ai
π Built an AI-powered Cricket Match Analysis platform using Googleβs Gemini 2.5 Flash and Google Cloud Platform.
The application analyzes cricket match videos/images and generates instant ball-by-ball insights with detailed statistics and visualizations.
Key Features
β
Shot type detection
β
Ball type identification
β
Ball speed estimation
β
Shot direction with wagon wheel visualization
β
Scoring zone radar charts
β
Player statistics (Strike Rate, Economy, etc.)
β
AI-generated ball-by-ball commentary
Tech Stack / GCP Services
βοΈ Gemini 2.5 Flash β Multimodal AI analysis for cricket footage
βοΈ Google Cloud Run β Serverless application deployment
βοΈ Google Cloud Storage β Media storage layer
βοΈ Google Cloud Build β CI/CD automation pipeline
Highlights
π Real-time AI-driven cricket analytics
π Multimodal video + image understanding
π Scalable serverless architecture on GCP
π Cost-efficient deployment using free-tier services
Excited about the possibilities of combining Generative AI with Sports Analytics.
#AI #GenerativeAI #Cricket #SportsAnalytics #GoogleCloud #GCP #GeminiAI #CloudRun #MachineLearning #ComputerVision #DataEngineering #Python


