Skip to main navigation Skip to search Skip to main content

Hyperspherical Quantization: Toward Smaller and More Accurate Models

  • Dan Liu*
  • , Xi Chen
  • , Chen Ma
  • , Xue Liu
  • *Corresponding author for this work

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 32 - Refereed conference paper (with host publication)peer-review

Abstract

Model quantization enables the deployment of deep neural networks under resource-constrained devices. Vector quantization aims at reducing the model size by indexing model weights with full-precision embeddings, i.e., codewords, while the index needs to be restored to 32-bit during computation. Binary and other low-precision quantization methods can reduce the model size up to 32×, however, at the cost of a considerable accuracy drop. In this paper, we propose an efficient framework for ternary quantization to produce smaller and more accurate compressed models. By integrating hyperspherical learning, pruning and reinitialization, our proposed Hyperspherical Quantization (HQ) method reduces the cosine distance between the full-precision and ternary weights, thus reducing the bias of the straight-through gradient estimator during ternary quantization. Compared with existing work at similar compression levels (~30×, ~40×), our method significantly improves the test accuracy and reduces the model size. © 2023 IEEE.
Original languageEnglish
Title of host publicationProceedings - 2023 IEEE Winter Conference on Applications of Computer Vision (WACV 2023)
PublisherIEEE
Pages5251-5261
ISBN (Electronic)978-1-6654-9346-8
DOIs
Publication statusPublished - 2023
Event23rd IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2023) - Waikoloa, United States
Duration: 3 Jan 20237 Jan 2023

Publication series

NameProceedings - IEEE Winter Conference on Applications of Computer Vision, WACV

Conference

Conference23rd IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2023)
PlaceUnited States
CityWaikoloa
Period3/01/237/01/23

Bibliographical note

Full text of this publication does not contain sufficient affiliation information. With consent from the author(s) concerned, the Research Unit(s) information for this record is based on the existing academic department affiliation of the author(s).

Research Keywords

  • Algorithms: Machine learning architectures
  • and algorithms (including transfer)
  • formulations
  • Image recognition and understanding (object detection, categorization, segmentation, scene modeling, visual reasoning)

Fingerprint

Dive into the research topics of 'Hyperspherical Quantization: Toward Smaller and More Accurate Models'. Together they form a unique fingerprint.

Cite this