Speech recognition in edge environments: an exploration of support and impact of model compression.
Automatic Speech Recognition (ASR) is an essential component of Human Computer Interaction. With the surge in mobile and home assistant IoT devices, there is a rising need for edge computing solutions supporting ASR applications. Whisper is a state-of-art ASR model that exhibits superior performance...
| Publicado en: | Language Resources & Evaluation Vol. 60; no. 1; pp. 1 - 21 |
|---|---|
| Autores principales: | , , , |
| Formato: | Artículo |
| Publicado: |
Springer Nature
Mar2026
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=192212482&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 192212482 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 1574020X 179V jtl: Language Resources & Evaluation issn: 1574020X maglogo: N pubinfo: dt: Mar2026 vid: 60 iid: 1 pid: 237 pub: Springer Nature artinfo: ui: 192212482 10.1007/s10579-025-09885-6 ppf: 1 ppct: 20 formats: tig: atl: Speech recognition in edge environments: an exploration of support and impact of model compression. aug: au: Joy, Carolene Martin, John Paul Joseph, Christina Terese Madhavan, Manu affil: https://ror.org/00gcgw028 Government College of Engineering Kannur, Kannur, India Indian Institute of Information Technology Kottayam, Kottayam, India su: Automatic speech recognition Edge computing Internet of things Integer approximations Computer memory management sug: subj: Automatic speech recognition Edge computing Internet of things Integer approximations Computer memory management keyword: Edge device Information and Computing Sciences Artificial Intelligence and Image Processing Quantization Speech recognition ab: Automatic Speech Recognition (ASR) is an essential component of Human Computer Interaction. With the surge in mobile and home assistant IoT devices, there is a rising need for edge computing solutions supporting ASR applications. Whisper is a state-of-art ASR model that exhibits superior performance. Being resource hungry, this model fails to deliver similar performance on resource constrained devices. Quantization is a thoroughly studied model compression technique in the premise of Deep Learning, which reduces the model size by using smaller integer representations for weights and activations. In this work, quantization is applied to the Whisper releases which are then deployed on Raspberry Pi and Jetson Orin Nano. To get a fair understanding, the models are also deployed on an AWS EC2 instance as well as on a personal computing device. This work is one of the initial attempts to study the performance of quantized Whisper models on edge devices. The quantized models achieved compression ratios of over 3.7 with only a marginal degradation in transcription accuracy. Real-time performance improved significantly, with up to 15% reduction in inference latency observed for the Base model on Raspberry Pi 5. Furthermore, quantized models demonstrated a notable reduction in memory usage, enabling deployment on devices with limited computational resources. pubtype: Academic Journal doctype: Article src: R language: English refInfo: copyright: @attributes: flag: Y dt: @attributes: year: 2026 holdings: @attributes: islocal: N |
|---|