A joint-loss approach for speech enhancement via single-channel neural network and MVDR beamformer

• Main Authors: Tan, Zhi-Wei, Nguyen, Anh Hai Trieu, Tran, Linh T. T., Khong, Andy Wai Hoong
• Other Authors: School of Electrical and Electronic Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2021
• Subjects: Engineering; Neural Beamforming; Complex Spectral Mapping
• Online Access: https://hdl.handle.net/10356/146260
• Institution: Nanyang Technological University

Recent developments of noise reduction involves the use of neural beamforming. While some success is achieved, these algorithms rely solely on the gain of the beamformer to enhance the noisy signals. We propose a framework that comprises two stages where the first-stage neural network aims to achiev...