深度学习目标检测综述.docx
需积分: 50 129 浏览量
2021-06-18
16:35:50
上传
评论 3
收藏 1.46MB DOCX 举报
目标检测方法发展综述
目标检测方法发展综述
摘要:目标检测是这样一种任务：在图像或者视频中，找到目标位置并进行分类。近年来，由于其广泛应
用而引起了重视。这篇文章综述了近年来基于深度学习的目标检测的最新进展，包括传统目标检测，一阶
段目标检测以及两阶段目标检测的分析，还分析了在识别任务中使用的一些突出的骨干网络和目标检测中
常用的数据集。比较好的目标检测网络对于具有实时性的目标检测研究更为必要。
关键词: 目标检测，发展综述，神经网络，数据集，轻量级网络
Review of development of target detection methods
Abstract: Target detection is a kind of task: in the image or video, find the target position and classification.
In recent years, it has attracted much attention because of its wide application. This paper reviews the latest
progress of target detection based on deep learning in recent years, including the analysis of traditional target
detection, one-stage target detection and two-stage target detection, and also analyzes some outstanding backbone
networks used in recognition tasks. At the same time, some contemporary lightweight classification networks for
edge devices are also provided. These lightweight networks are more necessary for real - time target detection
research.
Key words: target detection, development review, neural network, data set, lightweight
network
1
剩余11页未读，继续阅读
评论0
最新资源