Constrained Convolutional-Recurrent Networks to Improve Speech Quality with Low Impact on Recognition Accuracy
For a speech-enhancement algorithm, it is highly desirable to simultaneously improve perceptual quality and recognition rate. Thanks to computational costs and model complexities, it is challenging to train a model that effectively optimizes both metrics at the same time. In this paper, we propose a method for speech enhancement that combines local and global contextual structures information through convolutional-recurrent neural networks that improves perceptual quality. At the same time, we introduce a new constraint on the objective function using a language model/decoder that limits the impact on recognition rate. Based on experiments conducted with real user data, we demonstrate that our new context-augmented machine-learning approach for speech enhancement improves PESQ and WER by an additional 24.5% and 51.3%, respectively, when compared to the best-performing methods in the literature.
Code (0)
등록된 구현이 없습니다.
Tasks
DecoderLanguage ModelingLanguage ModellingSpeech EnhancementSimilar Papers 제목 키워드 기반
Single Channel Speech Enhancement Using Temporal Convolutional Recurrent Neural Networks
In recent decades, neural network based methods have significantly improved the performace of speech enhancement. Most of them estimate time-frequency (T-F) representation of target speech directly or indirectly, then re…
Speech EnhancementInplace Gated Convolutional Recurrent Neural Network For Dual-channel Speech Enhancement
For dual-channel speech enhancement, it is a promising idea to design an end-to-end model based on the traditional array signal processing guideline and the manifold space of multi-channel signals. We found that the idea…
Speech EnhancementA Lightweight Hybrid Dual Channel Speech Enhancement System under Low-SNR Conditions
Although deep learning based multi-channel speech enhancement has achieved significant advancements, its practical deployment is often limited by constrained computational resources, particularly in low signal-to-noise r…
Speech EnhancementJoint Blind Room Acoustic Characterization From Speech And Music Signals Using Convolutional Recurrent Neural Networks
Acoustic environment characterization opens doors for sound reproduction innovations, smart EQing, speech enhancement, hearing aids, and forensics. Reverberation time, clarity, and direct-to-reverberant ratio are acousti…
parameter estimationRoom Impulse Response (RIR)Speech EnhancementUltra-Low-Bitrate Speech Coding with Pretrained Transformers
Speech coding facilitates the transmission of speech over low-bandwidth networks with minimal distortion. Neural-network based speech codecs have recently demonstrated significant improvements in quality over traditional…
DecoderInductive Bias