Abstract:Nonnegative matrix factorization- (NMF-) based noise reduction methods can effectively improve the performance of environmental sound recognition. However, when the environmental sound overlaps highly with the noise, the spectral line loss and noise residue will occur in the low signal-to-noise ratio (SNR) condition. An adaptive noise reduction algorithm was proposed in this paper. First, noisy environmental sound is separated into estimated noise and environmental sound using NMF. Then, the estimated noise is used to calculate sound presence probability (SPP), which is adapted to decrease spectral line loss and achieve accurately estimated noise. Subsequently, the estimated noise combines with noisy environmental sound to obtain the estimated environmental sound. Finally, SPP is applied to reduce residual noise in the estimated environmental sound and reconstruct the environmental sound. The simulation results demonstrate that the proposed algorithm outperforms the traditional algorithms and NMF-based methods in terms of perceptual evaluation of speech quality (PESQ) and global SNR with increase of X% and X%, respectively. Moreover, the proposed method can effectively improve the environmental sound recognition rate. Particularly, the proposed method makes a 16.2% increase of F1-score in car horn recognition under the realistic acoustic condition.

A speech recognition system in high noise background based on discriminative learning of environmental features

A Robust Speech Recognition System in Noise Environment

Robust Speech Recognition Method Based on Discriminative Learning of Environmental Features

Robust Speech Recognition Method Based on Discriminative Environment Feature Extraction

Deep Learning for Environmentally Robust Speech Recognition: An Overview of Recent Developments

Discriminative environment feature extraction for robust speech recognition

Hierarchical speech recognition model in multi-noise environment

Study on method and system implementation of isolated word recognition system for noisy environment

Speech Recognition Algorithm in Complex Noisy Environments Based on Multi-Space Compensation

Speaker Recognition Algorithm Based on Sparse Constraint Discriminant Analysis

An Overview of Speech Feature Enhancement Method

Application of Progressive Learning Speech Enhancement Approach to Speech Recognition

Speech Recognition System Based on CDHMM/SOFMNN in Noisy Environment

A Noise Robust Front End Algorithm for Mandarin Speech Recognition and Performance Analysis

An Environment Adaptation Method for Robust Speech Recognition

Noisy speech recognition performance of discriminative HMMs

A Novel Speech/Noise Discrimination Method for Embedded ASR System

Noisy training for deep neural networks in speech recognition

Speech Recognition System Based on SCHMM/ANN in Noisy Environment

Adaptive Noise Reduction Algorithm Based on SPP and NMF for Environmental Sound Event Recognition under Low-SNR Conditions

Discriminative Learning of Additive Noise and Channel Distortions for Robust Speech Recognition