摘要
视频预测一直以来都是计算机视觉领域的热点问题,由于其广泛的实用价值和理论价值,引起了研究人员的广泛关注。该文对主流的视频帧预测算法进行了研究,首先介绍了视频预测领域的常见问题,并由基本架构对视频预测算法进行了分类,接下来介绍了该领域常用的数据集并给予评价,最后从视频预测算法的运行流程和最新的论文角度上,总结该领域算法的发展方向。
Video prediction has always been a key issue in the computer vision field.Because of its extensive pratical and theoraticalvalues,the video prediction technique has attracted attentions of many researchers.This paper reviews the main video prediction algorithms.Firstly,this paper introduce the common problems in video prediction,and classify many famous video prediction algorithms according to basic architectures.Then,the common datasets of video prediction are introduced and evaluated.Finally,on thebasis of the process of the algorithms and the latest papers,this paper summarize the development of algorithms in this field.
作者
刘志超
LIU Zhi-chao(School of Computer Science and Information Engineering,Hubei University,Wuhan 430062,China)
出处
《电脑知识与技术》
2021年第10期249-251,265,共4页
Computer Knowledge and Technology
关键词
视频帧预测
长短时记忆网络
自编码器
生成对抗网络
注意力机制
强化学习
video prediction
long short-term memory network
autoencoder
generative adversarial net
attention mechanism
rein⁃forcement learning