Object Detection in 20 Years A Survey
目标检测20年综述
ABSTRACT
Object detection, as of one the most fundamental and challenging problems in computer vision, has received great attention in recent years. Over the past two decades, we have seen a rapid technological evolution of object detection and its profound impact on the entire computer vision field. If we consider today’s object detection technique as a revolution driven by deep learning, then, back in the 1990s, we would see the ingenious thinking and long-term perspective design of early computer vision. This article extensively reviews this fastmoving research field in the light of technical evolution, spanning over a quarter-century’s time (from the 1990s to 2022). A number of topics have been covered in this article, including the milestone detectors in history, detection datasets, metrics, fundamental building blocks of the detection system, speedup techniques, and recent state-of-the-art detection methods.目标检测作为计算机视觉中最基本、最具挑战性的问题之一，近年来受到了广泛的关注。在过去的二十年里，我们看到了物体检测技术的快速发展及其对整个计算机视觉领域的深远影响。如果我们将今天的目标检测技术视为由深度学习驱动的革命，那么，回到20世纪90年代，我们将看到早期计算机视觉的巧妙思考和长期视角设计。本文广泛回顾了这一快速发展的研究领域的技术发展，跨越了四分之一个世纪的时间（从20世纪90年代到2022年）。本文涵盖了许多主题，包括历史上的里程碑式检测器、检测数据集、指标、检测系统的基本构建块、加速技术以及最新的最先进的检测方法。
KEYWORDS
Computer vision; convolutional neural networks (CNNs); deep learning; object detection; technical evolution.计算机视觉;卷积神经网络（CNN）;深度学习;物体检测;技术发展。
I. INTRODUCTION
Object detection is an important computer vision task that deals with detecting instances of visual objects of a certain class (such as humans, animals, or cars) in digital images. The goal of object detection is to develop computational models and techniques that provide one of the most basic pieces of knowledge needed by computer vision applications: What objects are where? The two most significant metrics for object detection are accuracy (including classification accuracy and localization accuracy) and speed.对象检测是一项重要的计算机视觉任务，它涉及检测数字图像中某类视觉对象（如人类、动物或汽车）的实例。目标检测的目标是开发计算模型和技术，提供计算机视觉应用所需的最基本知识之一：什么物体在哪里？目标检测的两个最重要的指标是准确性（包括分类准确性和定位准确性）和速度。
Object detection serves as a basis for many other computer vision tasks, such as instance segmentation [1], [2], [3], [4], image captioning [5], [6], [7], and object tracking [8]. In recent years, the rapid development of deep learning techniques [9] has greatly promoted the progress of object detection, leading to remarkable breakthroughs and propelling it to a research hot-spot with unprecedented attention. Object detection has now been widely used in many real-world applications, such as autonomous driving, robot vision, and video surveillance. Fig. 1 shows the growing number of publications that are associated with “object detection” over the past two decades.对象检测是许多其他计算机视觉任务的基础，例如实例分割[1]，[2]，[3]，[4]，图像字幕[5]，[6]，[7]和对象跟踪[8]。近年来，深度学习技术的快速发展[9]极大地推动了目标检测的进步，取得了显著的突破，并将其推向了前所未有的研究热点。目标检测已经广泛应用于自动驾驶、机器人视觉、视频监控等领域。图1显示了在过去二十年中与“目标检测”相关的出版物数量的增长。
As different detection tasks have totally different objectives and constraints, their difficulties may vary from each other. In addition to some common challenges in other computer vision tasks, such as objects under different viewpoints, illuminations, and intraclass variations, the challenges in object detection include, but are not limited to, the following aspects: object rotation and scale changes (e.g., small objects), accurate object localization, dense and occluded object detection, speedup of detection, and so on. In Section IV, we will give a more detailed analysis of these topics.由于不同的检测任务有着完全不同的目标和约束条件，其难度也会有所不同。除了其他计算机视觉任务中的一些常见挑战，例如不同视点下的对象，照明和类内变化，对象检测中的挑战包括但不限于以下方面：对象旋转和尺度变化（例如，小目标）、精确目标定位、密集和遮挡目标检测、检测加速等。在第四节中，我们将给予更详细的分析。
This survey seeks to provide novices with a complete grasp of object detection technology from many viewpoints, with an emphasis on its evolution. The key features are threefolds: a comprehensive review in the light of technical evolutions, an in-depth exploration of the key technologies and the recent state of the arts, and a comprehensive analysis of detection speedup techniques. The main clue focuses on the past, present, and future, complemented with some other necessary components in object detection, such as datasets, metrics, and acceleration techniques. Standing on the technical highway, this survey aims to present the evolution of related technologies, allowing readers to grasp the essential concepts and find potential future directions, while neglecting their technical specifics.本调查旨在从多个角度为新手提供对象检测技术的完整掌握，重点是其演变。主要特点有三个方面：根据技术发展进行全面回顾，深入探讨关键技术和最新技术状况，并全面分析检测加速技术。主要线索集中在过去，现在和未来，辅以对象检测中的其他必要组件，如数据集，度量和加速技术。本调查站在技术高速公路上，旨在呈现相关技术的演变，让读者掌握基本概念并找到潜在的未来方向，同时忽略其技术细节。
GPT解读
这篇论文《Object Detection in 20 Years: A Survey》全面回顾了过去二十年对象检测技术的发展历程，强调了其在计算机视觉领域的重要性和挑战性。以下是论文的核心内容概述：
对象检测的重要性：对象检测是计算机视觉中的基础任务，涉及在数字图像中检测特定类别的视觉对象实例。它为计算机视觉应用提供了基本的知识：对象在哪里。
技术演进：论文详细回顾了从1990年代到2022年的技术发展，包括里程碑式的检测器、数据集、评估指标、检测系统的基本构建块、加速技术以及最新的检测方法。
里程碑检测器：介绍了历史上的重要检测器，如Viola-Jones检测器、HOG检测器、DPM、R-CNN系列、YOLO系列、SSD、RetinaNet、CornerNet和DETR等。
数据集与评估指标：讨论了多个著名的对象检测数据集，如PASCAL VOC、ILSVRC、MS-COCO等，以及评估对象检测性能的指标，如平均精度（AP）和交并比（IoU）。
技术发展：分析了多尺度检测、上下文引导、硬负样本挖掘、损失函数和非极大值抑制（NMS）等关键技术的演进。
加速技术：探讨了提高检测器速度的方法，包括特征图共享计算、级联检测、网络剪枝和量化、轻量级网络设计以及数值加速。
最新进展：回顾了近年来的一些最新技术，如超越滑动窗口检测、旋转和尺度变化的鲁棒检测、更好的背景网络、改进的定位方法、与分割损失的学习、对抗性训练、弱监督对象检测和领域适应。
未来方向：提出了未来研究的可能方向，包括轻量级对象检测、端到端对象检测、小对象检测、3D对象检测、视频中的检测、跨模态检测以及开放世界检测。
作者信息：论文由Zhengxia Zou、Keyan Chen、Zhenwei Shi、Yuhong Guo和Jieping Ye共同撰写，他们在计算机视觉、模式识别和机器学习领域有着丰富的研究和实践经验。
这篇论文为读者提供了对象检测技术的全面视角，强调了其在计算机视觉领域的演变，并对未来的研究方向提供了深入的分析。