Weakly Guided Beamforming

Jakob Kienegger, Tal Peer, Sina Khanagha, Timo Gerkmann

This site acts as supplementary material to our work on Weakly Guided and Autoregressive Beamformer Parameterization for Generalizable Moving Speaker Extraction in Higher-Order Ambisonics . We present speech enhancement examples from a synthetic two-speaker dataset as well as recordings conducted in our acoustic lab.

Choose a dataset: Synthetic Dataset Recorded Dataset