CHE06 - AI-AURA (AI powered Acoustic Understanding, Recognition and Alerting)

Develop a GUI based interactive tool powered with AI that enables users to upload raw hydrophone recordings. The system will be further preprocess and extracts relevant acoustic features. We will develop machine learning (ML) models that are capable of learning and recognizing  underwater sound signatures and inputs the extracted features to detect and classify the signatures through the integration of model with the GUI. This will give an automated classification of signals to acoustic events like biophonic, anthropogenic or geophonic based on the training provided to ML model. The output will be presented in an intuitive way with confidence level and audio playback capable of hearing the sound by the user. In addition, the platform supports models to retrain and refine using new datasets to improve accuracy and adaptability.

A digital interactive tool with AI powered

Data analyst, AI/ML Knowledge, Experience in Signal processing, Oceanography and Ambient noise data (desirable)