Evaluation of Predictability for Speedometer using Deep Learning Models Based on Predictive Coding Theory
予測符号化理論に基づく深層学習モデルを用いた車載スピードメータの予測性評価
- Delivery
- Available on this site
- Format
- Price
- Non-members (tax incl.):¥1,100 Members (tax incl.):¥880
- Publication code
- 20216144
- Paper/Info type
- Proceedings (Autumn)
No.112-21
- Pages
- 1-6(Total 6 p)
- Date of publication
- Oct 2021
- Publisher
- JSAE
- Language
- Japanese
- Event
- 2021 JSAE Annual Congress (Autumn) [Online]
Detailed Information
Author(J) | 1) 竹本 敦, 2) 寺田 哲也, 3) 岡﨑 俊実, 4) 森 裕紀, 5) 尾形 哲也 |
---|---|
Author(E) | 1) Atsushi Takemoto, 2) Tetsuya Terada, 3) Toshimi Okazaki, 4) Yuki Mori, 5) Tetsuya Ogata |
Affiliation(J) | 1) マツダ, 2) マツダ, 3) マツダ, 4) 早稲田大学, 5) 早稲田大学 |
Affiliation(E) | 1) Mazda, 2) Mazda, 3) Mazda, 4) Waseda University, 5) Waseda University |
Abstract(J) | 車載表示機の視認性を定量的に評価すべく,本研究では大脳皮質の基本動作原理とされる予測符号化理論をもとに構築された深層学習モデルを用いてスピードメーターの動きの予測性評価を行った.その結果,学習データの特性やスピードメーターの仕様により予測性に差が見られため,結果より得られた知見について報告する. Translation |
Abstract(E) | Currently, there is a glowing need for the development of in-vehicle displays with high visibility, and for a method of quantitatively evaluating their visibility performance. In this research, we evaluated the predictability for speedometer needle movement using deep learning models based on the predictive coding theory. As a result, there were some differences in predictability depending on the characteristics of the learning data and the specifications of the in-vehicle meter, so we report the findings obtained from the results. |