Object Detection in 20 Years: A Survey
By Zhengxia Zou, Keyan Chen, Zhenwei Shi, Yuhong Guo, Jieping Ye. 2019
3,480 citations recorded by the source corpus.
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 fast-moving 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.
Subfields: Computer Vision and Pattern Recognition, Industrial and Manufacturing Engineering
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Trevor Darrell
- Xiaogang Wang
- Ming-Hsuan Yang
- Jitendra Malik
- Li Fei-Fei
- Andrew Zisserman
In Industrial and Manufacturing Engineering
- Yong Jae Lee
- Sergey Levine
- Jianfeng Gao
- Karol Hausman
- Gong Cheng
- Muhammad Hussain
- Lei Zhang
- Andy Zeng
Citations in the corpus
- Inner Monologue: Embodied Reasoning through Planning with Language Models (2022, background): “One common type of such feedback is object recognition [89, 90, 91, 92] that returns a list of”
- A Simple Framework for Open-Vocabulary Segmentation and Detection (2023, background): “Detection and segmentation have been long-standing problems in the vision community [8, 9, 37, 61].”
- SOCRATES: Introducing Depth in Visual Wildlife Monitoring Using Stereo Vision (2022, background): “State-of-the-art approaches to object detection are usually learning-based [12].”
- YOLOv1 to v8: Unveiling Each Variant–A Comprehensive Review of YOLO (2024, background): “From navigating through occlusions to efficiently managing scale variations and fine-grained object detection scenarios [9], YOLO faces a myriad of obstacles.”
- RSPrompter: Learning to Prompt for Remote Sensing Instance Segmentation Based on Visual Foundation Model (2023, background): “This process yields crucial information regarding the location (where), category (what), and shape (how) of each object [1], [2], [3], [4], [5], [6], [7], [8], [9].”
- YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception (2025, neutral): “Since the advent of deep convolutional neural networks [2], real-time object detection has rapidly evolved from multi-stage pipelines exemplified by the R-CNN series [4], [5], [23], [24] to highly optimized single-stage frameworks epitomized by the YOLO models [25].”
- YOLOv11: An Overview of the Key Architectural Enhancements (2024, background): “A crucial aspect of this domain is object detection[2], which involves the precise identification and localization of objects within images or video streams[3].”
- YOLOv4-5D: An Effective and Efficient Object Detector for Autonomous Driving (2021, background): “However, the accuracy of YOLOv4, especially for small objects, is still lower than a two-stage detector using a region proposal stage [37].”
- Computing Systems for Autonomous Driving: State of the Art and Challenges (2020, background): “through two typical phases: 1) conventional object detection phase and 2) deep learning supported object detection phase [67].”
- Segment Everything Everywhere All at Once (2023, background): “puter vision, as evidenced by the extensive literature on the subject [12, 11, 66, 37].”
- Towards Large-Scale Small Object Detection: Survey and Benchmarks (2022, background): “4, the size range of objects in SODA-D mainly comes to [10 , 30] and for SODA-A, it is strikingly [5 , 15] .”
- WebVoyager: Building an End-to-End Web Agent with Large Multimodal Models (2024, background): “Then the context will be updated and this interaction process continues until the model generates a terminating action or the maximum step is reached.”