S Devi, S Shoba, K Asutosh… - Circuits, Systems, and …, 2022 - search.proquest.com
The most commonly used training target is masking-based approach which maps noisy
speech to the time–frequency (T–F) unit and has a remarkable impact on the performance in …