Research and examination on implementation of super-resolution models using deep learning with INT8 precision

Shota Hirose, Naoki Wada, Jiro Katto, Heming Sun

研究成果: Conference contribution

2 被引用数 (Scopus)

抄録

Fixed-point arithmetic is a technique for treating weights and intermediate values as integers in deep learning. Since deep learning models generally store each weight as a 32-bit floating-point value, storing by 8-bit integers can reduce the size of the model. In addition, memory usage can be reduced, and inference can be much faster by hardware acceleration when special hardware for int8 inference is provided. On the other hand, when inferences are carried out by fixed-point weights, accuracy of the model is reduced due to loss of dynamic range of the weights and intermediate layer values. For this reason, inference frameworks such as TensorRT and TensorFlow Lite, provide a function called "calibration"to suppress the deterioration of the accuracy caused by quantization by measuring the distribution of input data and numerical values in the intermediate layer when quantization is performed. In this paper, after quantizing a pre-trained model that performs super-resolution, speed and accuracy are measured using TensorRT. As a result, the trade-off between the runtime and the accuracy is confirmed. The effect of calibration is also confirmed.

本文言語English
ホスト出版物のタイトル4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022 - Proceedings
出版社Institute of Electrical and Electronics Engineers Inc.
ページ133-137
ページ数5
ISBN(電子版)9781665458184
DOI
出版ステータスPublished - 2022
イベント4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022 - Jeju lsland, Korea, Republic of
継続期間: 2022 2月 212022 2月 24

出版物シリーズ

名前4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022 - Proceedings

Conference

Conference4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022
国/地域Korea, Republic of
CityJeju lsland
Period22/2/2122/2/24

ASJC Scopus subject areas

  • 人工知能
  • コンピュータ サイエンスの応用
  • 電子工学および電気工学

フィンガープリント

「Research and examination on implementation of super-resolution models using deep learning with INT8 precision」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