Encoder-Decoder Neural Architecture Optimization for Keyword Spotting

2021/06/04 by Tong Mo, Bang Liu, Mo, Tong +1
Computer Science · #FOS: Computer and information sciences #Machine Learning (cs.LG) #Multimedia (cs.MM) #Music and Audio Processing #Natural Language Processing Techniques #Speech Recognition and Synthesis

paper · pdf · doi:10.48550/arxiv.2106.02738

openalex publication_date 2021/06/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Keyword spotting aims to identify specific keyword audio utterances. In recent years, deep convolutional neural networks have been widely utilized in keyword spotting systems. However, their model architectures are mainly based on off-the shelfbackbones such as VGG-Net or ResNet, instead of specially designed for the task. In this paper, we utilize neural architecture search to design convolutional neural network models that can boost the performance of keyword spotting while maintaining an acceptable memory footprint. Specifically, we search the model operators and their connections in a specific search space with Encoder-Decoder neural architecture optimization. Extensive evaluations on Google's Speech Commands Dataset show that the model architecture searched by our approach achieves a state-of-the-art accuracy of over 97%.

Citations

• Neural Architecture Search with Reinforcement Learning
• Hello Edge: Keyword Spotting on Microcontrollers
• SNAS: Stochastic Neural Architecture Search
• Temporal Convolution for Real-time Keyword Spotting on Mobile Devices

Related

• End-to-end Keyword Spotting using Neural Architecture Search and Quantization
• Resource-efficient DNNs for Keyword Spotting using Neural Architecture Search and Quantization
• AutoKWS: Keyword Spotting with Differentiable Architecture Search
• Predicting detection filters for small footprint open-vocabulary keyword\n spotting
• Stochastic Adaptive Neural Architecture Search for Keyword Spotting