A HEARING DEVICE COMPRISING A NOISE REDUCTION SYSTEM

The present application relates to hearing aid configured to be worn by a user at or in an ear, the hearing aid comprising an input unit for providing at least one electric input signal (IN(k,m)) in a time frequency representation k, m, where k and m are frequency and time indices, respectively, and k represents a frequency channel, the at least one electric input signal being representative of sound and comprising target signal components and noise components; and a signal processor comprising an SNR estimator (SNR-EST) for providing a target signal-to-noise ratio (SNR) estimate for said at least one electric input signal in said time frequency representation (k,m); an SNR-to-gain converter (SNR2G) for converting said target signal-to-noise ratio estimates to respective gain values (G(k,m)) in said time frequency representation (k,m), where said SNR-to-gain converter (SNR2G) comprises a neural network (NN), wherein the weights of the neural network have been trained with a plurality of training signals (SNRp(k,m), G-OPTp(k,m)); and wherein said SNR-to-gain converter (SNR2G) is configured to provide a maximum amount of noise reduction.