Implementing Keyword Spotting on the MCUX947 Microcontroller with Integrated NPU

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jakuš, Petar, Džapo, Hrvoje
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912441309331456
author Jakuš, Petar
Džapo, Hrvoje
author_facet Jakuš, Petar
Džapo, Hrvoje
contents This paper presents a keyword spotting (KWS) system implemented on the NXP MCXN947 microcontroller with an integrated Neural Processing Unit (NPU), enabling real-time voice interaction on resource-constrained devices. The system combines MFCC feature extraction with a CNN classifier, optimized using Quantization Aware Training to reduce model size with minimal accuracy drop. Experimental results demonstrate a 59x speedup in inference time when leveraging the NPU compared to CPU-only execution, achieving 97.06% accuracy with a model size of 30.58 KB, demonstrating the feasibility of efficient, low-power voice interfaces on embedded platforms.
format Preprint
id arxiv_https___arxiv_org_abs_2506_08911
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Implementing Keyword Spotting on the MCUX947 Microcontroller with Integrated NPU
Jakuš, Petar
Džapo, Hrvoje
Human-Computer Interaction
Hardware Architecture
Machine Learning
Performance
Sound
This paper presents a keyword spotting (KWS) system implemented on the NXP MCXN947 microcontroller with an integrated Neural Processing Unit (NPU), enabling real-time voice interaction on resource-constrained devices. The system combines MFCC feature extraction with a CNN classifier, optimized using Quantization Aware Training to reduce model size with minimal accuracy drop. Experimental results demonstrate a 59x speedup in inference time when leveraging the NPU compared to CPU-only execution, achieving 97.06% accuracy with a model size of 30.58 KB, demonstrating the feasibility of efficient, low-power voice interfaces on embedded platforms.
title Implementing Keyword Spotting on the MCUX947 Microcontroller with Integrated NPU
topic Human-Computer Interaction
Hardware Architecture
Machine Learning
Performance
Sound
url https://arxiv.org/abs/2506.08911