Project Title AI-Powered Audio Analysis Software for Sleep Respiratory Detection & Snoring Classification Project Description I am looking for an expert in Audio Signal Processing and Machine Learning to develop a specialized software tool for analyzing 8-hour sleep audio recordings. The system must process audio data to identify, categorize, and log specific respiratory events using provided training datasets. 1. Detection & Classification Requirements The software must accurately detect and quantify the following: • Normal Breathing: Total event count. • Labored Breathing: Total event count. • Snoring Intensity Levels: o Low Snoring: 10 – 30 dB o Moderate Snoring: 30 – 50 dB o Heavy Snoring: 50 – 80 dB o Explosive Snoring: > 80 dB • Apneic Events: o Apnea: Total cessation of breath for more than 10 seconds. o Hypopnea: At least 50% reduction in breathing amplitude. 2. Features & Analytics • Manual Validation Module: An interactive interface that allows the user to manually review, confirm, or reject detected events (Manual Override) for data verification. • Statistical Reporting: Automated calculation of average duration and maximum duration for both Apnea and Hypopnea events. • Automated Visualization: A dashboard that generates a color-coded Pie Chart showing the percentage (%) distribution of all detected categories. 3. Data & Training I will provide an extensive, high-quality dataset containing labeled samples of all categories (Snoring levels, Apnea, Normal, and Labored breathing) for model training and testing.