摘要
目标检测是计算机视觉领域中一项十分重要的任务,指的是对图像或视频中的目标进行分类和定位,因其被广泛应用于视频检测、目标跟踪等任务中而受到关注。本文首先介绍了目标检测中经常使用的基准数据集和评估指标,同时列举说明了重要的主干网络架构和流行的目标检测算法。最后,我们比较了这些体系结构在多个指标上的性能,对所述算法进行了总结,并对未来前景进行了展望。
Object detection is a very important task in the field of computer vision,which refers to the classification and location of objects in images or videos.It has attracted much attention because it is widely used in video detection,object tracking and other tasks.This paper first introduces the benchmark datasets and evaluation indicators often used in object detection,and lists the important backbone network architecture and popular object detection algorithms.Finally,we compare the performance of these architectures on multiple indicators,summarize the algorithms,and look forward to the future.
作者
钱承武
QIAN Cheng-wu(Faculty of Electrical Engineering and Computer Science,Ningbo University,Ningbo 315211,China)
出处
《无线通信技术》
2022年第4期24-29,共6页
Wireless Communication Technology
关键词
目标检测
深度学习
卷积神经网络
object detection
deep learning
convolutional neural network(CNN)
作者简介
钱承武(1998-),男,宁波大学信息科学与工程学院硕士研究生,主要研究方向为人工智能、计算机视觉。
