深層学習用活性化関数によるニューラルネットワーク選点法トモグラフィの改良
深層学習用活性化関数によるニューラルネットワーク選点法トモグラフィの改良
カテゴリ: 論文誌(論文単位)
グループ名: 【C】電子・情報・システム部門
発行日: 2023/07/01
タイトル(英語): Tomographic Image Reconstruction by Neural Network Colocation Method - Improvement by using Deep Learning type Activation Functions
著者名: 寺西 大(広島工業大学)
著者名(英語): Masaru Teranishi (Hiroshima Institute of Technology)
キーワード: トモグラフィ,ニューラルネットワーク,画像再構成,関数近似,深層学習,活性化関数 tomography,neural networks,image reconstruction,function approximation,deep learning,activation function
要約(英語): Neural Network Collocation Method (NNCM) is a kind of model fitting method and is an effective imaging technique for application plasma tomographic imaging field that has limited to the number of views. The reconstructed image is represented as nonlinear combination of the basis functions by the activation functions of NNCM. NNCM could fit models based on few spatial location data because every basis function cover whole reconstruction region and are trained based on error back propagation algorithm. Therefore, the image could be effectively reconstructed even from few-view poor projection data. However, application of NNCM to practical imaging has an issue that NNCM requires many training epochs even to reconstruct a small-scale image. The main reason of the issue is attributed to that the small gradient value of sigmoid activation functions let the training speed slow.The paper proposes an improvement of NNCM by replacing the activation functions of hidden layers of NNCM with deep learning type ones which have larger gradient values, ReLU and tanh, to increase training speed of NNCM. Numerical simulation results show the effectiveness of the proposed method.
本誌: 電気学会論文誌C(電子・情報・システム部門誌) Vol.143 No.7 (2023) 特集:2022年電子・情報・システム部門大会
本誌掲載ページ: 694-700 p
原稿種別: 論文/日本語
電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/143/7/143_694/_article/-char/ja/
受取状況を読み込めませんでした
