PROSE: Perceptual Risk Optimization for Speech Enhancement

by   Jishnu Sadasivan, et al.
indian institute of science

The goal in speech enhancement is to obtain an estimate of clean speech starting from the noisy signal by minimizing a chosen distortion measure, which results in an estimate that depends on the unknown clean signal or its statistics. Since access to such prior knowledge is limited or not possible in practice, one has to estimate the clean signal statistics. In this paper, we develop a new risk minimization framework for speech enhancement, in which, one optimizes an unbiased estimate of the distortion/risk instead of the actual risk. The estimated risk is expressed solely as a function of the noisy observations. We consider several perceptually relevant distortion measures and develop corresponding unbiased estimates under realistic assumptions on the noise distribution and a priori signal-to-noise ratio (SNR). Minimizing the risk estimates gives rise to the corresponding denoisers, which are nonlinear functions of the a posteriori SNR. Perceptual evaluation of speech quality (PESQ), average segmental SNR (SSNR) computations, and listening tests show that the proposed risk optimization approach employing Itakura-Saito and weighted hyperbolic cosine distortions gives better performance than the other distortion measures. For SNRs greater than 5 dB, the proposed approach gives superior denoising performance over the benchmark techniques based on the Wiener filter, log-MMSE minimization, and Bayesian nonnegative matrix factorization.


page 11

page 12


GMM based multi-stage Wiener filtering for low SNR speech enhancement

This paper proposes a single-channel speech enhancement method to reduce...

Normalized Features for Improving the Generalization of DNN Based Speech Enhancement

Enhancing noisy speech is an important task to restore its quality and t...

Improving the Generalizability of Deep Neural Network Based Speech Enhancement

Enhancing noisy speech is an important task to restore its quality and t...

Low Noise Non-Linear Equalization Using Neural Networks and Belief Propagation

Nonlinearities can be introduced into communication systems by the physi...

UNetGAN: A Robust Speech Enhancement Approach in Time Domain for Extremely Low Signal-to-noise Ratio Condition

Speech enhancement at extremely low signal-to-noise ratio (SNR) conditio...

Speech enhancement using ego-noise references with a microphone array embedded in an unmanned aerial vehicle

A method is proposed for performing speech enhancement using ego-noise r...

Please sign up or login with your details

Forgot password? Click here to reset