Resource-Efficient Speech Mask Estimation for Multi-Channel Speech Enhancement
arXiv:2007.11477
Abstract
While machine learning techniques are traditionally resource intensive, we are currently witnessing an increased interest in hardware and energy efficient approaches. This need for resource-efficient machine learning is primarily driven by the demand for embedded systems and their usage in ubiquitous computing and IoT applications. In this article, we provide a resource-efficient approach for multi-channel speech enhancement based on Deep Neural Networks (DNNs). In particular, we use reduced-precision DNNs for estimating a speech mask from noisy, multi-channel microphone observations. This speech mask is used to obtain either the Minimum Variance Distortionless Response (MVDR) or Generalized Eigenvalue (GEV) beamformer. In the extreme case of binary weights and reduced precision activations, a significant reduction of execution time and memory footprint is possible while still obtaining an audio quality almost on par to single-precision DNNs and a slightly larger Word Error Rate (WER) for single speaker scenarios using the WSJ0 speech corpus.
References in corpus (13)
- Distilling the Knowledge in a Neural Network
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients
- Binarized Neural Networks
- Ternary Weight Networks
- Pyroomacoustics: A Python package for audio room simulations and array processing algorithms
- Training deep neural networks with low precision multiplications
- Training Deep Neural Networks with 8-bit Floating Point Numbers
- Soft Weight-Sharing for Neural Network Compression
- Recurrent Neural Networks With Limited Numerical Precision
- Streamlined Deployment for Quantized Neural Networks
- Resource-Efficient Neural Networks for Embedded Systems
- ProdSumNet: reducing model parameters in deep neural networks via product-of-sums matrix decompositions