A neural attention model for speech command recognition

TL;DR

This paper presents a convolutional recurrent neural network with attention for speech command recognition, achieving state-of-the-art accuracy while maintaining a small model size, and providing interpretability of the audio regions influencing decisions.

Contribution

The paper introduces a novel attention-based neural network architecture that improves speech command recognition accuracy and interpretability over previous models.

Findings

Achieved 94.1% accuracy on Google Speech Commands V1

Achieved 94.5% accuracy on V2 for 20 commands

Model has only 202K parameters

Abstract

This paper introduces a convolutional recurrent network with attention for speech command recognition. Attention models are powerful tools to improve performance on natural language, image captioning and speech tasks. The proposed model establishes a new state-of-the-art accuracy of 94.1% on Google Speech Commands dataset V1 and 94.5% on V2 (for the 20-commands recognition task), while still keeping a small footprint of only 202K trainable parameters. Results are compared with previous convolutional implementations on 5 different tasks (20 commands recognition (V1 and V2), 12 commands recognition (V1), 35 word recognition (V1) and left-right (V1)). We show detailed performance results and demonstrate that the proposed attention mechanism not only improves performance but also allows inspecting what regions of the audio were taken into consideration by the network when outputting a given category.