Self-Attention (SA)-ConvLSTM Encoder–Decoder Structure-Based Video Prediction for Dynamic Motion Estimation

Video prediction, which is the task of predicting future video frames based on past observations, remains a challenging problem because of the complexity and high dimensionality of spatiotemporal dynamics. To address the problems associated with spatiotemporal prediction, which is an important decis...

Full description

Bibliographic Details

Published in:	Applied sciences Vol. 14; no. 23; p. 11315
Main Authors:	Kim, Jeongdae, Choo, Hyunseung, Jeong, Jongpil
Format:	Journal Article
Language:	English
Published:	Basel MDPI AG 01.12.2024
Subjects:	Computational linguistics Datasets encoder–decoder Forecasts and trends Language processing Natural language interfaces Neural networks Performance evaluation SA-ConvLSTM self-attention memory module spatiotemporal video prediction
ISSN:	2076-3417, 2076-3417
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!