Structural sparsification for Far-field Speaker Recognition with GNA
arXiv:1910.11488
Abstract
Recently, deep neural networks (DNN) have been widely used in speaker recognition area. In order to achieve fast response time and high accuracy, the requirements for hardware resources increase rapidly. However, as the speaker recognition application is often implemented on mobile devices, it is necessary to maintain a low computational cost while keeping high accuracy in far-field condition. In this paper, we apply structural sparsification on time-delay neural networks (TDNN) to remove redundant structures and accelerate the execution. On our targeted hardware, our model can remove 60% of parameters and only slightly increasing equal error rate (EER) by 0.18% while our structural sparse model can achieve more than 1.5x speedup.
submitted to icassp2020
References in corpus (7)
- VoxCeleb2: Deep Speaker Recognition
- VoxCeleb: a large-scale speaker identification dataset
- Pyroomacoustics: A Python package for audio room simulations and array processing algorithms
- Deep Speaker: an End-to-End Neural Speaker Embedding System
- Exploring Sparsity in Recurrent Neural Networks
- Learning Intrinsic Sparse Structures within Long Short-Term Memory
- Voices Obscured in Complex Environmental Settings (VOICES) corpus