Kohei Saijo

independent vector analysis (2) source separation (2) speech enhancement (2) unsupervised source separation (2) unsupervised learning (2) model architecture (1) deep learning (1) speech dereverberation (1) audio source separation (1) autoregressive model (1) end-to-end training (1) teacher-student framework (1) direction of arrival (1) siamese network (1) neural beamformer (1) permutation invariant training (1) model scalability (1) compute budget (1) unsupervised speech separation (1) reverberation as supervision (1)