STREAMING SPEECH TO TEXT ON ANDROID A SOCKET.IO BASED SERVER APPROACH FOR ANDROID MOBILE APPLICATION

Douglas Rakasiwi Nugroho (Universitas Bina Nusantara)
Christopher Limawan (Universitas Bina Nusantara)
Kelvin (Universitas Bina Nusantara)

Algebra : Jurnal Pendidikan Sosial dan Sains
Vol. 5 No. 4 (2025): Algebra : Jurnal Pendidikan Sosial dan Sains

This paper details a robust system enabling real-time Speech-to-Text capabilities on Android devices, leveraging a Socket.IO-based server architecture to manage audio streams and integrate with advanced language models. This approach effectively addresses the inherent challenges of on-device processing, such as latency, power consumption, and computational overhead, by offloading the intensive Speech-to-Text and Natural Language Processing tasks to a scalable server infrastructure. This distributed processing paradigm ensures minimal resource drain on the client device while maximizing accuracy and responsiveness.