Deep Learning Framework Estimates Acoustic Contexts from First-Order Ambisonic Speech Recordings
A deep learning framework using spectro-spatial covariance vectors from first-order Ambisonics achieves over fifty percent error reduction in acoustic parameter ...

