EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Collaborative Multi-Compression Acceleration Mechanism for Neural Networks in Keyword Spotting

Junbang Jiang, Rui Pu, Jin Li, Man Zhu

Symmetry · 18.08.2026 · 18 (8) · 1387

DOI: 10.3390/sym18081387 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Symmetry, online erschienen am 18.08.2026. Verfasst von Junbang Jiang und 3 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Hubei University of Technology, Hubei Provincial Department of Education, Cooperation Agreement on the Project of Design and Process Manufacturing Technology of 3D Stacked Code Flash Memory Chip.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To address the large model size, high computational cost, and limited deployment resources of keyword spotting models on edge platforms, this study proposes a collaborative multi-compression framework for lightweight deployment. Built on LiteKWS-Net, an attention-enhanced 2-D convolutional backbone, the framework combines adaptive importance-aware structured pruning, mixed-precision quantization, and quantization-aware multi-stage knowledge distillation. The retrained teacher reaches 97.90% (mean, 100,813 parameters, 0.385 MiB). MPDQ reaches 95.53 ± 1.16% at 8.27× theoretical weight compression. AIASP reaches 97.59% at a 30% target and 43.9% realized sparsity. The final joint model reaches 96.82% and, under ideal packed sparse mixed-precision storage, has a 51.55× theoretical weight-compression factor relative to the FP32 teacher; sparse-index overhead is excluded. On a Jetson Nano, the TensorRT FP16 network-body benchmark reports 2.86 ms latency and 0.69 mJ per inference.

Förderung

  • Hubei University of Technology
  • Hubei Provincial Department of Education
  • Cooperation Agreement on the Project of Design and Process Manufacturing Technology of 3D Stacked Code Flash Memory Chip

Zitieren

APA

Jiang, J., Pu, R., Li, J., & Zhu, M. (2026). A Collaborative Multi-Compression Acceleration Mechanism for Neural Networks in Keyword Spotting. Symmetry, 18(8), 1387. https://doi.org/10.3390/sym18081387

Vancouver

Jiang J, Pu R, Li J, Zhu M. A Collaborative Multi-Compression Acceleration Mechanism for Neural Networks in Keyword Spotting. Symmetry. 2026;18(8):1387. doi:10.3390/sym18081387

BibTeX
@article{Jiang2026,
  author = {Jiang, Junbang and Pu, Rui and Li, Jin and Zhu, Man},
  title = {A Collaborative Multi-Compression Acceleration Mechanism for Neural Networks in Keyword Spotting},
  journal = {Symmetry},
  year = {2026},
  volume = {18},
  number = {8},
  pages = {1387},
  doi = {10.3390/sym18081387}
}
Kostenlos starten →