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

Heming Sun, Zhengxue Cheng, Masaru Takeuchi, Jiro Katto

研究成果: Conference contribution

2 被引用数 (Scopus)

抄録

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.

本文言語English
ホスト出版物のタイトル2020 IEEE International Conference on Image Processing, ICIP 2020 - Proceedings
出版社IEEE Computer Society
ページ3359-3363
ページ数5
ISBN(電子版)9781728163956
DOI
出版ステータスPublished - 2020 10月
イベント2020 IEEE International Conference on Image Processing, ICIP 2020 - Virtual, Abu Dhabi, United Arab Emirates
継続期間: 2020 9月 252020 9月 28

出版物シリーズ

名前Proceedings - International Conference on Image Processing, ICIP
2020-October
ISSN(印刷版)1522-4880

Conference

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

ASJC Scopus subject areas

  • ソフトウェア
  • コンピュータ ビジョンおよびパターン認識
  • 信号処理

フィンガープリント

「End-To-End Learned Image Compression with Fixed Point Weight Quantization」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル