End-To-End Learned Image Compression with Fixed Point Weight Quantization

Heming Sun, Zhengxue Cheng, Masaru Takeuchi, Jiro Katto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Learned image compression (LIC) has reached the traditional hand-crafted methods such as JPEG2000 and BPG in terms of the coding gain. However, the large model size of the network prohibits the usage of LIC on resource-limited embedded systems. This paper presents a LIC with 8-bit fixed-point weights. First, we quantize the weights in groups and propose a non-linear memory-free codebook. Second, we explore the optimal grouping and quantization scheme. Finally, we develop a novel weight clipping fine tuning scheme. Experimental results illustrate that the coding loss caused by the quantization is small, while around 75% model size can be reduced compared with the 32-bit floating-point anchor. As far as we know, this is the first work to explore and evaluate the LIC fully with fixed-point weights, and our proposed quantized LIC is able to outperform BPG in terms of MS-SSIM.

Original languageEnglish
Title of host publication2020 IEEE International Conference on Image Processing, ICIP 2020 - Proceedings
PublisherIEEE Computer Society
Pages3359-3363
Number of pages5
ISBN (Electronic)9781728163956
DOIs
Publication statusPublished - 2020 Oct
Event2020 IEEE International Conference on Image Processing, ICIP 2020 - Virtual, Abu Dhabi, United Arab Emirates
Duration: 2020 Sept 252020 Sept 28

Publication series

NameProceedings - International Conference on Image Processing, ICIP
Volume2020-October
ISSN (Print)1522-4880

Conference

Conference2020 IEEE International Conference on Image Processing, ICIP 2020
Country/TerritoryUnited Arab Emirates
CityVirtual, Abu Dhabi
Period20/9/2520/9/28

Keywords

  • Image compression
  • fine-tuning
  • fixed-point
  • neural networks
  • quantization

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Signal Processing

Fingerprint

Dive into the research topics of 'End-To-End Learned Image Compression with Fixed Point Weight Quantization'. Together they form a unique fingerprint.

Cite this