aTENNuate based masked channel signal reconstruction in simulated environments

Developed an SSM architecture to predict masked channel audio signal at a desired location in a simulated room given source and microphone configuration. This model learns the spatial audio representation via self-supervised learning technique