主成分分析とMSSA法を用いた動画予測手法の技術改良
英字タイトル:
Technological Improvement of Movie Prediction Method by PCA and MSSA
著者:
上赤 一馬 Kazuma KAMIAKA 出町 和之 Demachi KAZUYUKI 榊原 洋志 Hiroshi SAKAKIBARA 笠原 直人 Naoto KASAHARA
発刊日:
公開日:
Image predictionMSSA(Multi-channel Singular Spectrum Analysis)PCA(Principal Component Analysis)
Technological Improvement of Movie Prediction Method by PCA and MSSA
著者:
上赤 一馬 Kazuma KAMIAKA 出町 和之 Demachi KAZUYUKI 榊原 洋志 Hiroshi SAKAKIBARA 笠原 直人 Naoto KASAHARA
発刊日:
公開日:
カテゴリ: 第10回
キーワードタグ:Image predictionMSSA(Multi-channel Singular Spectrum Analysis)PCA(Principal Component Analysis)
概要
Abstract In recent years, there are so many systems using movie. So, we propose a new method to predict the near-future images using movie. To predict the movi e, the Principal Component Analysis (PCA) and Multi-Channel Singular Spectrum Analysis (MSSA) are used in this method. Predicted movie obtained from this metho d is almost the same as the original movie, but improvement is required to reduce the calculation time and to increase the precision of the prediction for real -time image prediction. So, in this research, we improve the algorithm and managed to reduce the calculation time and increase the precision of the prediction.
