Speech Phonation Microservice

Built a microservice using FastAPI to analyze speech samples for testing Parkinson’s disease. The service utilizes Librosa and Parselmouth for audio analysis, extracting features such as pitch, jitter, shimmer, and HNR. The results are stored in a Firebase database for easy access and retrieval. The microservice is containerized using Docker, ensuring easy deployment and scalability. It’s part of a larger project made for H2AI - 25 targeting the Neuro AI Innovation challenge. This microservice was containerized using Docker and deployed on Amazon ECS. The architecture is designed to be scalable and efficient, allowing for easy integration with other services and applications.

Features

• • Speech Analysis: The microservice analyzes speech samples to extract key features relevant to Parkinson’s disease detection, including pitch, jitter, shimmer, and HNR.
• • FastAPI Framework: Built using FastAPI, the microservice provides a robust and efficient API for handling requests and responses, ensuring high performance and low latency.
• • Audio Processing Libraries: Utilizes Librosa and Parselmouth for advanced audio processing and feature extraction, enabling accurate analysis of speech samples.
• • Firebase Integration: The results of the analysis are stored in a Firebase database, providing a scalable and secure solution for data storage and retrieval.
• • Docker Containerization: The microservice is containerized using Docker, allowing for easy deployment and scalability across different environments.
• • Amazon ECS Deployment: Deployed on Amazon ECS, ensuring high availability and scalability, making it suitable for production use.