sensifai
Audio Event Recognition
Real-time, zero-dependency AI engine for recognizing and timestamping 6 critical acoustic event categories.
关于此模型
Audio Event Recognition (AER) is an ultra-fast, zero-dependency AI engine designed for real-time acoustic sound processing. Operating 100% client-side inside web browsers via native Web Audio API, it extracts 80-dimensional MFCC and Delta-MFCC features to detect 6 critical sound categories with 97.3% precision: AMBIENT, SHOUTING, CRYING, IMPACT, ALARM_SIREN, and EXPLOSION. Fully integrated with Logto OIDC for authentication and Gatekeeper REST API for pay-as-you-go metered billing.
核心功能
Real-time Audio Detection
Zero-Dependency Inference
6 Acoustic Sound Classes
Delta-MFCC Feature Extraction
Export CSV & JSON Reports
使用场景
- Distress & Emergency Monitoring
- Acoustic Surveillance
- Smart Environment Security
定价计划
