MusicNet

Tesis: Tesis de pregrado

Resumen

This project addresses the challenge of real-time musical training over unstable networks, where latency and packet loss degrade user experience and hinder learning. We developed MusicNet, a browser-based multiplayer piano training platform that combines gamification, pitch detection, and a neural network-based packet loss concealment model (PARCnet). The system allows users to practice
scales with synchronized feedback, even under adverse network conditions. By integrating WebRTC audio transmission and predictive reconstruction, MusicNet delivers an interactive and resilient learning
experience, validated through performance metrics and user feedback.
Fecha de lectura19 may. 2025
Idioma originalEspañol
Institución de lectura
  • Pontificia Universidad Javeriana

Palabras clave

  • Retardo en Comunicaciones

Citar esto

'