Object Detection in 20 Years: A Survey
Open accessThe DOI matches its Crossref record; 309 of the 309 references it cites were found in a registry or an index; 84 of the 192 sources its claims cite were checked for support.
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 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.
Overview
CoChat Summary
TL;DR
This survey reviews two decades of object detection research (1990s–2022), tracing the transition from handcrafted-feature detectors to deep-learning-based two-stage, one-stage, and transformer detectors. It organizes the field around milestone detectors, datasets and metrics, key building blocks (multi-scale detection, context, hard negative mining, loss functions, NMS), speed-up techniques, and recent advances.
Key findings
- The field splits into a "traditional object detection period (before 2014)" and a "deep learning based detection period (after 2014)"; RCNN raised VOC07 mAP from 33.7% (DPM-v5) to 58.5%, SPPNet reached VOC07 mAP=59.2% at >20× faster than R-CNN, Fast RCNN reached 70.0% mAP at >200× faster than R-CNN, and Faster RCNN reached COCO mAP@.5=42.7%, VOC07 mAP=73.2% at 17fps.
- One-stage detectors traded some accuracy for speed: YOLO ran at 155fps with VOC07 mAP=52.7% (enhanced version 45fps, 63.4%), SSD reached COCO mAP@.5=46.5% at 59fps, and RetinaNet's focal loss brought one-stage accuracy to COCO mAP@.5=59.1%.
- Keypoint- and transformer-based detectors pushed accuracy further: CornerNet COCO mAP@.5=57.8%, CenterNet 61.1%, and Deformable DETR 71.9% on MS-COCO; the paper notes the top-10 COCO detection methods are all transformer-based.
- Datasets scaled dramatically: VOC07 has 5k train images/12k objects, VOC12 11k/27k, MS-COCO-17 164k images and 897k annotated objects over 80 categories, and Open Images 1,910k images with 15,440k boxes over 600 categories. Evaluation moved from miss-rate vs. FPPW to FPPI, then to AP/mAP, with MS-COCO averaging over IoU thresholds 0.5–0.95.
Methods
- Design: a narrative literature survey spanning roughly a quarter-century (1990s to 2022), structured around technical evolution rather than a systematic/quantitative review protocol.
- Coverage: milestone detectors (VJ, HOG, DPM, RCNN, SPPNet, Fast/Faster RCNN, FPN, YOLO, SSD, RetinaNet, CornerNet, CenterNet, DETR), datasets and metrics, technical building blocks, speed-up techniques, and recent advances (rotation/scale robustness, backbones, localization, segmentation loss, adversarial training, weakly supervised detection, domain adaptation).
- Analysis: qualitative synthesis of reported detector accuracies and speeds, with figures/tables summarizing dataset statistics and accuracy trends; no new experiments are conducted.
Limitations
- The paper states no explicit limitations section; it frames its scope as presenting the evolution of technologies "while neglecting their technical specifics," and notes that the evolution order shown in figures is primarily to assist reader understanding and "may have temporal overlap."
- Evident from the text: as a survey it reports no original experiments or controlled benchmarks, so the cited mAP/speed figures come from the original papers and are not independently verified; the coverage is also selective (milestone detectors and selected recent advances) rather than exhaustive.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
55 Valid · 18 Not valid · 32 Not checked · 50 Own results and interpretations
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“In the deep learning era, due to the increase of computing power, bootstrap was shortly discarded in object detection during 2014-2016 [16-20].”
What the sources say
- Not checked — no full text of the sourcecites [16] matched to “Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation” (2014) source
- Disputedcites [17] matched to “Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition” (2014) sourceExperimental — accuracy not yet measured
Section 4.1 (page unavailable) says the authors used standard hard negative mining, iterated once, to train their object-detection SVMs. That is contrary to citing this 2014 detector as evidence that bootstrapping had been discarded. The paper does not establish a 2014–2016 trend or attribute one to increased computing power. (LLM, conf=0.78)
Checked by a reasoning model — may be wrong.
- Not checked — no full text of the sourcecites [18] matched to “Fast R-CNN” (2015) source
- Partly supportedcites [19] matched to “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks” (2017) sourceExperimental — accuracy not yet measured
Section 4.2 (p. 10) says Fast R-CNN abandoned the SVM step that, in SPPnet, used hard-negative mining. This supports a narrower change in one detection system, but not the claim that bootstrapping was discarded across object detection during 2014–2016 or that increased computing power caused the change. (LLM, conf=0.83)
- Could not verifycites [20] matched to “You Only Look Once: Unified, Real-Time Object Detection” (2016) sourceExperimental — accuracy not yet measured
The Introduction (p. 1) contrasts YOLO with earlier detection methods, and Training (p. 3) describes how YOLO is trained. Neither passage addresses whether bootstrapping was discarded in object detection during 2014–2016 or attributes such a change to increased computing power; searches for bootstrapping and hard-negative mining found no passage that does. (LLM, conf=0.72)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Faster RCNN is the first near-realtime deep learning detector (COCO mAP@.5=42.7%, VOC07 mAP=73.2%, 17fps with ZF-Net [48]).”
What the sources say
- Partly supportedcites [19] matched to “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks” (2017) sourceExperimental — accuracy not yet measured
Section 4.2 reports 42.7% COCO mAP@0.5; Section 4.1 reports 73.2% VOC 2007 mAP and 17 fps with ZF-Net. These are results from different model configurations: the 73.2% VOC result uses VGG-16, not ZF-Net. The paper describes Faster R-CNN as a fast detector but does not establish that it was the first near-real-time deep-learning detector. (LLM, conf=0.88)
- Partly supportedcites [47] matched to “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks” (2017) sourceExperimental — accuracy not yet measured
Section 4.2 reports 42.7% COCO mAP@0.5, and Section 4.1 reports 73.2% VOC 2007 mAP and 17 fps with ZF-Net. The accuracy figures are from VGG-16 experiments, not the 17-fps ZF-Net configuration. The paper describes Faster R-CNN as a fast detector but does not establish that it was the first near-real-time deep-learning detector. (LLM, conf=0.88)
- Partly supportedcites [48] matched to “Visualizing and understanding convolutional networks” (2014) sourceExperimental — accuracy not yet measured
Section 3 (p. 4) describes the authors’ convolutional-network architecture, supporting the citation as an identification of the ZF-Net backbone. The paper does not report Faster R-CNN’s near-real-time status, COCO or VOC07 mAP, or 17-fps speed, so it does not support those performance claims. (LLM, conf=0.88)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In the case of sliding window based detectors, the imbalance between backgrounds and objects could be as extreme as 10 7 :1 [71].”
What the sources say
- Partly supportedcites [71] matched to “What Makes for Effective Detection Proposals?” (2016) sourceExperimental — accuracy not yet measured
The Introduction (p. 1) says sliding-window detection can search 10^6–10^7 windows per image. It does not say that 10^7 of those windows are backgrounds for each object, so it does not establish the claimed 10^7:1 imbalance. (LLM, conf=0.88)
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“It achieves state-of-the-art performance on MSCOCO dataset (COCO mAP@.5=71.9%).”
What the sources say
- Supportedcites [43] Zhu, 2020, Deformable DETR: Deformable transformers for end-to-end object detection, arXiv:2010.04159 source
Section 5.3 (pp. 8–9) compares Deformable DETR with state-of-the-art methods on COCO. Its table reports an AP50 of 71.9 for the ResNeXt-101 + DCN configuration with test-time augmentation; the accompanying text reports 52.3 overall AP for that configuration. (LLM, conf=0.91)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Using FPN in a basic Faster R-CNN system, it achieves state-of-the-art single model detection results on the COCO dataset without bells and whistles (COCO mAP@.5=59.1%).”
What the sources say
- Supportedcites [24] matched to “Feature Pyramid Networks for Object Detection” (2017) source
The Introduction describes a state-of-the-art single-model COCO result using FPN with a basic Faster R-CNN detector, without bells and whistles. Section 5.2.3 reports AP@0.5 of 59.1 on COCO test-dev and says the single-model entry surpassed competition winners. Page numbers are not provided in the copy. (LLM, conf=0.94)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
50 more claims · 21 with their cited sources checked
Create a free account to see all 155 claims and to verify the 3 references still open.
- It may state two separate claims as one. 40
- Probably not a substantial claim. 83
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“From navigating through occlusions to efficiently managing scale variations and fine-grained object detection scenarios [9], YOLO faces a myriad of obstacles.”
From YOLOv1 to v8: Unveiling Each Variant–A Comprehensive Review of YOLO (2024) · cites this paper as [9]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“The speed-up techniques in object detection can be divided into three levels of groups: speed up of 'detection pipeline', 'detector backbone', and 'numerical computation'. , as shown in Fig. 9. Refer to [123] for a more detailed version.”
From Object Detection in 20 Years: A Survey (2023) · cites this paper as [123]
The cited source says what the claim attributes to it.
In the citing paper
“The few surveys that mention both object detection and model compression [11] focus on the acceleration of”
From A survey of model compression strategies for object detection (2023) · cites this paper as [11]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“In addition to that, most of the reviews in related fields review the object detection model and model compression technology separately, and there is a lack of comprehensive reviews for both.”
From A survey of model compression strategies for object detection (2023) · cites this paper as [11]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ II. OBJECT DETECTION IN 20 YEARS
“Fig. 2: A road map of object detection. Milestone detectors in this figure: VJ Det. [10, 11], HOG Det. [12], DPM [1315], RCNN [16], SPPNet [17], Fast RCNN [18], Faster RCNN [19], YOLO [20-22], SSD [23], FPN [24], Retina-Net [25], CornerNet [26], CenterNet [27], DETR [28].”
§ Object detection accuracy improvements
“CenterNet: X. Zhou et al proposed CenterNet [40] in 2019.”
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 4×
“As the accuracy of a detector depends heavily on its feature extraction networks, in this paper, we refer to the backbone networks, e.g. the ResNet and VGG, as the “engine” of a detector.”
InfluentialImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 5×
“In 2012, the world saw the rebirth of convolutional neural networks (CNNs) [35].”
InfluentialLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472 · cited 10×
“The main idea of multi-reference detection [19, 22, 23, 41, 47, 51] is to first define a set of references (a.”
InfluentialFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 1×
“It contains a detection challenge using ImageNet images [61].”
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 3×
“Some well-known detection models of the early time: (a) Eigenfaces [95], (b) Shared weight networks [96], (c) Space displacement networks (Lenet-5) [97], (d) Haar wavelets of VJ detector [10].”
BackgroundMethodDistinctive Image Features from Scale-Invariant Keypoints
Lowe, D. G. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000029664.99615.94 · Cited by 55,589 · cited 1×
“HOG can be considered as an important improvement of the scale-invariant feature transform [33, 34] and shape contexts [35] of its time.”
BackgroundYou Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803 · cited 9×
“L2 loss is prevalent in early research [16], [20], [51], but it is highly affected by outliers and prone to gradient explosion.”
InfluentialDensely Connected Convolutional Networks
Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.243 · Cited by 48,566
Research works printed but not linked: 9
- 10.5555/3045118.3045336 · doi:10.5555/3045118.3045336
- Girshick, 2023, Discriminatively Trained Deformable Part Models, Release 5
- 10.1109/TPAMI.2016.2577031 · DOI · Paper page
- Krasin, 2017, OpenImages: A Public Dataset for Large-Scale Multi-Label and Multi-Class Image Classification
- Alexe, What is an object, 73
- Müller, When does label smoothing help, 32, 1
- 10.1109/CVPR.2016.308 · DOI · Paper page
- 10.5555/2969033.2969125 · doi:10.5555/2969033.2969125
- 10.1109/ICCV.2017.244 · DOI · Paper page
Cited by
Items 2,101–2,200 of 3,144
Object Detection by Network Based on CapsNet Architecture and Attention LayersPreprint
Norouzi, M., Hassanpour, H., & Sorkhi, A. G. (2023) · Research Square · doi:10.21203/rs.3.rs-3133189/v1 · Cited by 1
Noise-Tolerant Hardware-Aware Pruning for Deep Neural Networks
Lu, S., Chen, C., Zhang, K., Zheng, Y., Hu, Z., Hong, W., Li, G., & Yao, X. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-36625-3_11
Akinwehinmi, O., del Amo, P. A., & Solsona, F. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4496307
ANALYSIS OF INFORMATION TECHNOLOGIES FOR REMOTE IDENTIFICATION OF DYNAMIC OBJECTS
Nikulina, O., Severyn, V. P., Кондратов, О. М., & Rekova, N. Y. (2023) · Bulletin of National Technical University KhPI Series System Analysis Control and Information Technologies · doi:10.20998/2079-0023.2023.01.17 · Cited by 10
Evaluating an Artificial Intelligence Approach for Converting Sketches to UI Layouts
Dang, S. (2023) · Carleton University's Institutional Repository (MacOdrum Library, Carleton University) · doi:10.22215/etd/2023-15460 · Cited by 2
IEC-YOLOX:A Lightweight Object Detection Algorithm for UAV Target DetectionPreprint
Song, Q., Xu, S., Xue, Y., & Yu, J. (2023) · Research Square · doi:10.21203/rs.3.rs-3174559/v1
Platform for the Analysis of the Density of Shopping Carts in a Supermarket
Sandoval, J. C., Tostado, G. F., González, M. R., et al. (2023) · Advances in intelligent systems and computing · doi:10.1007/978-3-031-38344-1_33
Raw camera data object detectors: an optimisation for automotive processing and transmissionPreprint
Chan, P. H., Wei, C., Huggett, A., et al. (2023) · doi:10.36227/techrxiv.23807499
Otm-Fusion: An Image Preprocessing Method for Object Detection in Grayscale ImagePreprint
Wan, D., Lu, R., Hu, B., et al. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4532335
Feature Selector: an effective module for robust high-voltage switchgear detection
Chen, J., Xie, W. C., Zhang, L., et al. (2023) · International Conference on Electronic Information Technology (EIT) · doi:10.1117/12.2685486
“However, due to the limited availability of training data and the presence of disturbances such as fog, low light, and noise in outdoor environments, these models suffer from low generalization [2].”
BackgroundMachine Learning-based Auto-merge Program for Nine-directional Ocular Photography
Park, S. H., Lee, W. H., Kang, T. S., et al. (2023) · Journal of the Korean Ophthalmological Society · doi:10.3341/jkos.2023.64.8.734
Learning Point Processes and Convolutional Neural Networks for Object Detection in Satellite ImagesPreprint
Mabon, J., Ortner, M., & Zerubia, J. B. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4545325
Oh, S.-H., Park, H., & Park, J. (2023) · Journal of Animal Science and Technology · doi:10.5187/jast.2023.e81
“GoogLeNet [6] improved the inefficient structure, and could model deeper than VggNet by using an inception module that included a 1 × 1 convolution filter asymmetrically connected and layers that were not fully connected, resulting in a smaller model size and faster computation.”
BackgroundMethodDSWA: A Dilated Shift Window Attention Method for Chinese Named Entity Recognition (S)
Hou, X., Zhu, C., & Zhu, W. (2023) · Proceedings · doi:10.18293/dmsviva2023-106
Chen, S., Zhang, C., & Mu, H. (2023) · Research Square · doi:10.21203/rs.3.rs-3321326/v1
Correlation and Foreground Attention to Improve Object Detection
Dong, Y., Yue, X., Xu, Z., & Xie, S. (2023) · Proceedings - International Conference on Image Processing · doi:10.1109/icip49359.2023.10222828 · Cited by 1
“We use Average Precision (AP) and Average Recall (AR) [17] as evaluation metrics.”
MethodSTDG: Fast and Lightweight SNN Training Technique Using Spike Temporal LocalityPreprint
Cai, Z., Kalatehbali, H. R., Walters, B., et al. (2023) · doi:10.22541/au.169446658.81099759/v1
Computer Vision-Based Pointer-Style Mechanical Dial Reading Method for LNG Terminal Safety
Liu, H., Lyu, J., Liu, S., et al. (2023) · International Conference on Transportation Information and Safety · doi:10.1109/ictis60134.2023.10243772
“Deep learning-based target detection methods can be categorised into two main types, based on their algorithmic characteristics [9], as two-stage or one-stage target detection algorithms.”
BackgroundTriple Attention Vision Transformers for Remote Sensing Image Classification
Li, S., Guo, J., Li, J., et al. (2023) · IEEE International Conference on Electronic Information and Communication Technology (ICEICT) · doi:10.1109/iceict57916.2023.10245719
“Different from applications such as object detection [11] and semantic segmentation [12], scene classification focuses on global macro information, and generally tends to classify regions as a whole according to the semantic information of the scene.”
MethodJoint Edge-Guided and Spectral Transformation Network for Self-supervised X-Ray Image Restoration
Huang, S., Zou, W., Gao, H., Yang, W., Chen, H., Niu, S., Qi, T., & Ma, J. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-44210-0_33
Behavior Recognition Method Based on Object Detection for Power Operation Scenes
Liu, H., & Wang, L. (2023) · Lecture notes in electrical engineering · doi:10.1007/978-981-99-6187-0_48
Lu, L., Dai, F., & Brilakis, I. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4584805
Finding Antibodies in Cryo-EM densities with CrAIPreprint
Mallet, V., Rapisarda, C., Minoux, H., et al. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.09.27.559736
“In particular, we leverage the conserved structure of Abs as a prior information[14] to formulate our problem as a special instance of 3D object detection [49, 64].”
MethodYOLOv5-based Fatigued Driving Detection
Guan, L. (2023) · IEEE International Conference on Sensors, Electronics and Computer Engineering (ICSECE) · doi:10.1109/icsece58870.2023.10263342
A Streamlined Framework for Bev-Based 3d Object Detection with Prior MaskingPreprint
Tong, Q., Zhang, J., Yan, C. C., et al. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4597486
Real-Time Monitoring of Rectangular Indicator Lights in Intelligent Power Plants
Zhao, Z., Chen, L., Yang, X., et al. (2023) · International Conference on Robotics, Artificial Intelligence and Intelligent Control (RAIIC) · doi:10.1109/raiic59453.2023.10281018
Survey of object detection in terahertz images
Li, H. (2023) · Applied and Computational Engineering · doi:10.54254/2755-2721/19/20231016
An Adaptive Method for Dynamic Background Compensation Based on SIFT
Tao, X., Jia, Q., Du, T., et al. (2023) · IEEE Information Technology and Mechatronics Engineering Conference (ITOEC) · doi:10.1109/itoec57671.2023.10292043
“In order to achieve subsequent processing such as recognition, modeling, and analysis of multiple moving objects, fast and accurate detection and tracking of these objects has become an important issue[1].”
BackgroundWu, R., & Guo, D. (2023) · doi:10.1145/3615884.3629428
“While prevailing object detection tasks, such as pedestrian detection, traffic signal recognition, and remote sensing target identification [29], predominantly focus on foreground objects, the background elements within images often harbor vital information essential for a comprehensive scene…”
BackgroundEB-YOLOX:A Balanced One-Stage Object Detection Model Concentrating on Global FeaturesPreprint
Wang, Y., Wang, K., Liu, Z., & Luo, G. (2023) · Research Square · doi:10.21203/rs.3.rs-3519769/v1
Construction of a Novel Data Set for Pedestrian Tree Species Detection Using Google Street View Data
Ooi, M., Valdez, D. A. S., Rogers, M., et al. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-45382-3_28
Find Drivable Segments from Road Image using Depth and RGB Image
Li, X. (2023) · doi:10.5121/csit.2023.131921
“Real-time object detection and image segmentation have a considerable leap in their performance, benefiting from the escalation of GPU and deep learning performance [1].”
BackgroundNardi, E. E. D. S., Padilha, B., Kamaura, L. T., & Ferreira, J. E. (2023) · Journal of Information and Data Management · doi:10.5753/jidm.2023.3179 · Cited by 1
“Zou et al. provide a comprehensive analysis of research over the last 20 years in the field of object detection at [ Zou et al ., 2023], covering topics such as milestone detectors, detection datasets, metrics, and state-of-the-art detection meth-ods, among others.”
BackgroundDistributed Carbon Sink Detection Network for Mangrove Environment
Zhang, J., Wang, Y., Zhang, Y., et al. (2023) · IEEE International Conference on Pattern Recognition and Artificial Intelligence (PRAI) · doi:10.1109/prai59366.2023.10332110
“The single-stage detection, on the other hand, is characterized by end-to-end, which is about one-step, and therefore has a faster speed loss of some accuracy [6].”
BackgroundA Machine Learning Approach to Unusual Object Detection in Faint Satellite Imagery
Sreeja, C. K., & Ameenudeen, P. E. (2023) · International Conference on Smart Computing and Communications (ICSCC) · doi:10.1109/icscc59169.2023.10335060
Pixel representations, sampling, and label correction for semantic part detection
Huang, J.-C., Lin, Y.-L., & Fang, W.-C. (2023) · Machine Vision and Applications · doi:10.1007/s00138-023-01493-0
“The results indicate that Faster R-CNN, a two-stage method, is significantly slower than YOLOv8, aligning with documented constraints [65–67].”
Retrieval-Based Methodology for Few-Sample Logo Recognition
Moralis, D., Tzelepi, M., & Tefas, A. (2023) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp59012.2023.10337712
“Different detectors could also be used [14].”
TRAFFIC VIOLATION AND DRIVER ATTENTION BASED SCORE CALCULATION USING DEEP LEARNING
(2023) · International Research Journal of Modernization in Engineering Technology and Science · doi:10.56726/irjmets46750
Intelligent Driver Behavior and Emotion Detection System
Lau, C.-Y., Yuen, M.-C., Yung, C.-W., et al. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia231097
Multi Scale Forest Fire Small Target Detection Technology
Cao, Y., Zhou, J. J., Wang, G., et al. (2023) · IEEE International Conference on Civil Aviation Safety and Information Technology (ICCASIT) · doi:10.1109/iccasit58768.2023.10351746
“The YOLOv5 model consists of Input, backbone,neck, and head [10], and the network structure is shown in Figure 1.”
MethodFast Traffic Sign Detection for Two-Way Roads using Detachable Onboard Cameras
de Paula, M. B., & Jung, C. R. (2023) · SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI) · doi:10.1109/sibgrapi59091.2023.10347130
“To reduce running times or hardware requirements, some authors focused on developing very lightweight CNN backbones [1]–[3], hardware implementations for embedded systems [4], or scaling methods that try to adapt the size (depth/width) of CNNs based on available hardware [5], [6].”
BackgroundUDA-HOID: Unsupervised Domain Adaptation for Human-Object Interaction Detection
Jiang, M., Huang, W., Li, M., Meng, B., Ren, J., Bai, R., & Yang, Y. (2023) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai59109.2023.00095
“However, in computer vision tasks, we have a consensus that high-level features contain more semantic information [12].”
Background2023 Index Proceedings of the IEEE Vol.111
(2023) · Proceedings of the IEEE · doi:10.1109/jproc.2023.3345570
Research on Explosive Identification and Localization for Explosive Filling Robot
Zhu, Y., Li, H., Zuo, Z., et al. (2023) · International Conference on Communication, Image and Signal Processing (CCISP) · doi:10.1109/ccisp59915.2023.10355835
“Experimental tests confirm its effectiveness. development of YOLO has profoundly impacted the entire field of computer vision[11].”
BackgroundComputing 2D Skeleton via Generalized Electric Potential
Ma, G., Wang, X., Liu, X., et al. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-8549-4_29
Wu, L., Cao, Z., & Chen, Z. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-8549-4_28
MERGE: Multi-Entity Relational Reasoning Based Explanation in Visual Question Answering
Pan, M., Li, Q., & Tian, Q. (2023) · IEEE Intl Conf on Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive Intelligence and Computing, Intl Conf on Cloud and Big Data Computing, Intl Conf on Cyber Science and Technology Congress (DASC/PiCom/CBDCom/CyberSciTech) · doi:10.1109/dasc/picom/cbdcom/cy59711.2023.10361383
“Therefore, compared to traditional computer vision tasks such as object detection [1], VQA models need the ability to handle more complex image-text inference.”
BackgroundAlgifari, M. H., Ashari, I. F., Nugroho, E., et al. (2023) · Journal of Applied Informatics and Computing · doi:10.30871/jaic.v7i2.6761
“Apart from that, testing will also utilize several parameters, such as Mean Average Precision (mAP), Intersection over Union (IoU), Precision, and Recall, to evaluate the optimal level of object detection [15][16][17].”
MethodSiamese YOLO V5 with Structure coefficient for object-level change detectionPreprint
Zhang, Y., Pang, J., Li, B., et al. (2023) · Research Square · doi:10.21203/rs.3.rs-3806822/v1
Continual Object Detection with Deep Neural Networks
Menezes, A. G. (2023) · Digital Library of Theses and Dissertations (Universidade de São Paulo) · doi:10.11606/t.55.2023.tde-08012024-161234
Farklı Derin Sinir Ağı Modelleri Birleştirilerek Yakın Kızılötesi Görüntülerde Yüz Tespiti
UNLUHISARCIKLI, A., & Toprak, A. N. (2024) · Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi · doi:10.28948/ngumuh.1373458
Evaluation of Deep Learning Model for Detecting Electronic Components in Few-Shot Learning Scenarios
Rocha, A., Fernandes, B., & Honorato, L. (2023) · IEEE Latin American Conference on Computational Intelligence (LA-CCI) · doi:10.1109/la-cci58595.2023.10409312
Rong, H., Qian, M., Ma, T., Jin, D., & Sheng, V. S. (2024) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3643565 · Cited by 3
Object Detection under Finger Occlusion in AR Geography Assisted Teaching System by Using PCCNet
Yan, P., Chen, H., Dong, S., & Chen, X. (2023) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc53992.2023.10394387
“Several approaches have been proposed to address the occlusion in object detection field [18].”
BackgroundJauch, C., Leitritz, T., & Huber, M. F. (2023) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc53992.2023.10394319
“An in-depth review of object detection can be found in [6].”
BackgroundAn AI-based 450-kV Computed Tomography Solution for Hairpin Welding Inspection
Tschechne, S., Rothe, L., Kaťuch, P., et al. (2024) · e-Journal of Nondestructive Testing · doi:10.58286/29250
YOLOv7-CBAM for Similar Casting Inspection
Xu, Z., Wang, X., Shu, K., & Fan, J. (2023) · International Conference on New Trends in Computational Intelligence (NTCI) · doi:10.1109/ntci60157.2023.10403682
“The development of hardware and technology has made it possible to use object detection algorithms [4] and cameras to detect real-time video streams of industrial castings.”
BackgroundObject Detection Based on Priori Attribute Information Aggregation Method
Li, C., Pan, J., Liang, J., et al. (2024) · Lecture notes in electrical engineering · doi:10.1007/978-981-99-9243-0_36
Zakeri, E., Xie, W., & Babiker, I. (2023) · International Conference on Mechatronics and Machine Vision in Practice (M2VIP) · doi:10.1109/m2vip58386.2023.10413372
Multi-step Ahead Visual Trajectory Prediction for Object Tracking using Echo State Networks
Manibardo, E. L., Laña, I., & Ser, J. D. (2023) · IEEE International Conference on Intelligent Transportation Systems (ITSC) · doi:10.1109/itsc57777.2023.10422485
“In short, TbD architectures exploit the excellent performance now provided by object detectors [7].”
BackgroundDetection of Punjabi Newspaper Articles Using a Deep Learning Approach
Kumar, A., & Lehal, G. S. (2024) · Lecture notes in electrical engineering · doi:10.1007/978-981-99-8661-3_30
Technological prototype with artificial intelligence for military security in river environments
Sánchez, S., Casillo, D., Merlano, A., Barrera, J., Morales, A., & Contreras, E. (2023) · IEEE Colombian Caribbean Conference (C3) · doi:10.1109/c358072.2023.10436304
SISTEM INSPEKSI PERMUKAAN BAJA BERBASIS DEEP LEARNING MENGGUNAKAN METODE ANCHOR-FREE
Yuliyanto, S., Amani, N. F., Akhyar, F., & Usman, K. (2022) · Jurnal Ilmiah Teknik Mesin Elektro dan Komputer · doi:10.51903/juritek.v2i3.364 · Cited by 2
Niu, H., & Chen, Y. (2023) · Agriculture automation and control · doi:10.1007/978-3-031-52645-9_1
A Central Similarity Hashing Method via Weighted Partial-Softmax Loss
Li, M., Fu, Y., Li, Z., Zhang, D., & Wan, Z. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-0862-8_21
Tree Crown Detection using UAV-captured High-Resolution Aerial Images for Baghdad University Campus
Wadod, M. M., & Mohammed, F. G. (2024) · Journal Port Science Research · doi:10.36371/port.2023.special.11
Lightweight Neural Network Design for Real-time Human Tracking with Visual Modality
Limawan, C., & Soewito, B. (2024) · International Journal of Computing and Digital Systems · doi:10.12785/ijcds/150140
Development and Key Technologies of the Cognition system of Autonomous Driving
Zhang, Y., Feng, J., Huang, M., Liu, Q., & Fu, W. (2023) · Highlights in Science Engineering and Technology · doi:10.54097/hset.v56i.10835
Object Detection and Instance Segmentation in Construction SitesPreprint
Zhang, C. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4750768
Hsu, C.-Y., Lu, Y., & Jiang, Y.-L. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4750909
Using Neural Factorization of Shape and Reflectance for Ball Detection
Monté, X., van der Kaaij, J., van der Weerd, R., et al. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-55015-7_11
“The last 20 years have seen many breakthroughs in object detection techniques, well documented by several surveys [7,27,29].”
BackgroundExperience learning——an artificial intelligence framework suitable for exploring the unknownPreprint
He, Q., Wang, S., & Si, J. (2024) · Research Square · doi:10.21203/rs.3.rs-3984524/v1 · Cited by 1
Buffered Gaussian Modeling for Vectorized HD Map Construction
Shi, A., Chen, H., Lu, H., et al. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10445925
“Intersection over union (IoU) is widely used as a scale-invariant metric in object detection [11], which could resolve this problem.”
BackgroundApplication of Cascade Methods as a Universal Object Detection Tool
Matalov, D. P., Usilin, S. A., Nikolaev, D., & Arlazarov, V. V. (2023) · Pattern Recognition and Image Analysis · doi:10.1134/s1054661823040302 · Cited by 1
Learning point process models for vehicles detection using CNNs in satellite images
Mabon, J., Ortner, M., & Zerubia, J. B. (2023) · doi:10.1109/sitis61268.2023.00022
“While object detection has been thoroughly studied for the past 20 years [1], small object detection in optical satellite images remains challenging due to the limited spatial resolution, where objects of interest such as vehicles are only few pixels large, and thus lack visual information.”
Yuan, Y., Wang, J., & Li, W. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4767281
Research on Small Target Detection Algorithm Based on Improved YOLOv5s
Xiumei, Z., & Chen, B. (2024) · Lecture notes in electrical engineering · doi:10.1007/978-981-99-7545-7_35
Tang, F., & Zhang, Q. (2024) · International Journal of Computer Science and Information Technology · doi:10.62051/ijcsit.v2n1.20
“Specifically, workers are required to manually verify the positioning of pipe sections prior to their attachment to or detachment from the drill string, a step critical for the subsequent actions of the iron roughneck[4].”
BackgroundExtracting semantic information of cycling lane based on street view image
Lei, X., & Meng, N. (2024) · International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID) · doi:10.1117/12.3026479
“YOLO [18] , SSD [19] and RetinaNet [20] series algorithm are typical representatives of one-stage algorithm.”
MethodDiverse Feature-Level Guidance Adjustments for Unsupervised Domain Adaptative Object Detection
Zhu, Y., Liu, C., Bai, Y., Wang, C., Wei, C., Li, Z., & Zhou, Y. (2024) · Applied Sciences · doi:10.3390/app14072844 · Cited by 1
Zhang, H., Zhang, Y., Huang, Q., Han, Y., & Zhao, M. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4781952
Adaptive Spatial Perception Network
Cai, W., Tan, S., Gong, K., & Yan, W. (2023) · International Conference on Intelligent Management and Software Engineering (IMSE) · doi:10.1109/imse61332.2023.00020
Oriented Cell Dataset: efficient imagery analyses using angular representationPreprint
Kirsten, L. N., Angonezi, A., Oliveira, F. R., Faccioni, J. L., Cassel, C. B., de Sousa, D. S., Vedovatto, S., Jung, C. R., & Lenz, G. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.04.05.588327 · Cited by 1
“General-purpose object detection with HBBs is already a well-consolidated problem in computer vision with several improvements in recent years, as noted in the recent survey paper [9].”
BackgroundInfluentialEdgeAware YOLO: Enhancing Small Object Edge Detection in Base Station Equipment Inspection
Zhang, X., Lü, T., Wang, J., Xu, A., & Xu, J. (2023) · International Conference on Artificial Intelligence Innovation (ICAII) · doi:10.1109/icaii59460.2023.10497409 · Cited by 1
Cultured Fish Length Estimation Based on Keypoint and Binocular Stereo VisionPreprint
Wei, S., Yu, H., Zhang, X., et al. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4795485
Systems for Processing Unstructured Multimedia Information
Pukhalskyi, B., & Kyriienko, O. (2024) · NaUKMA Research Papers Computer Science · doi:10.18523/2617-3808.2023.6.62-68
Ding, Z., Jin, N., Zhou, Y., et al. (2024) · PFG – Journal of Photogrammetry Remote Sensing and Geoinformation Science · doi:10.1007/s41064-024-00285-z
“These methods can efficiently process large-scale and high-dimensional data, and use neural network models to learn the characteristics of landslide images and spatiotemporal data, effectively improving the accuracy and efficiency of landslide recognition (Zou et al. 2023).”
Multi-scale feature fusion and lightweight decoupled Head in small object detectionPreprint
Lv, G., Zhang, Z., Yuan, M., Ding, Z., & Ma, G. (2024) · Research Square · doi:10.21203/rs.3.rs-4301366/v1
GlassOnly: Transparent object dataset for object detection
Mai, X. (2024) · Applied and Computational Engineering · doi:10.54254/2755-2721/57/20241306
Small Object Detection via a Dense Connection and Feature Enhancement Network
Qiang, B., Chen, L., Guo, S., Liu, L., & Zhang, S. (2023) · IEEE International Conference on Computer and Communications (ICCC) · doi:10.1109/iccc59590.2023.10507587
“Among those object detection methods based on deep learning, two categories of which are always receiving close attention, two-stage and one-stage detection method [2].”
MethodPronina, O. I., & ПОЛІЩУК, А. А. (2024) · Вісник Херсонського національного технічного університету · doi:10.35546/kntu2078-4481.2024.1.39
Nuclear Track Quantification in Cr-39 Detectors Using Deep LearningPreprint
Mohamed, S., Alahmari, S. S., Khan, A., et al. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4816632
Adaptive Fusion of Global Pooling Feature Attention for Object Detection
He, T., Yang, Z., Li, X., et al. (2023) · International Conference on Internet of Things, Automation and Artificial Intelligence (IoTAAI) · doi:10.1145/3653081.3653119
Park, C., Kim, W., Park, M. S., et al. (2024) · Journal of the Korean Society for Precision Engineering · doi:10.7736/jkspe.023.149
ASLAN, T., & Yağımlı, M. (2024) · Gazi Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi · doi:10.17341/gazimmfd.1306981
DCTE-LLIE: A Dual Color-and-Texture-Enhancement-Based Method for Low-Light Image Enhancement
Wang, H., Cao, J., Yang, L., & Huang, J. (2024) · Computers · doi:10.3390/computers13060134
ПОРІВНЯННЯ МЕТОДІВ ВИЯВЛЕННЯ ОБ’ЄКТІВ В КОМП’ЮТЕРНОМУ ЗОРІ
ЦИВАДИЦЬ, П., СКРИПНИК, Т., & ВОЗНЮК, Л. (2024) · Herald of Khmelnytskyi National University Technical sciences · doi:10.31891/2307-5732-2024-333-2-41
An Incremental Few Shot Approach for Litter Detection
Liao, Y., Gao, W., Mao, S., & Zhang, S. (2024) · Asia Conference on Algorithms, Computing and Machine Learning (CACML) · doi:10.1145/3654823.3654869
“However, classical object detection methods, such as the two-stage detectors and the one-stage detectors [31], rely on a large amount of training data and rich computation resource to achieve good detection results.”
BackgroundResearch on learner classroom behavior recognition based on YOLOv5 algorithm
Han, K., & Xu, J. (2023) · International Conference on Information Education and Artificial Intelligence (ICIEAI) · doi:10.1145/3660043.3660075
“Many video-based human behavior recognition algorithms have been published both domestically and internationally [6], and corresponding technologies are continuously advancing, leading to an annual increase in recognition accuracy.”
BackgroundRecognition and analysis system of steel stamping character based on machine vision
Wenming, C., Tianhong, T., Dongtai, L., et al. (2024) · Automatika · doi:10.1080/00051144.2024.2362575
“Among these, object detection methods [7] have gained significant traction in engineering endeavours due to their broad applicability and straightforward implementation.”
Metadata
- Authors
- Zou, Z., Chen, K., Shi, Z., Guo, Y., & Ye, J.
- Affiliations
- Shanghai Artificial Intelligence Laboratory (CN)
- Carleton University (CA)
- Alibaba Group (China) (CN)
- Venue
- Proceedings of the IEEE
- Published
- 2023-03
- Volume
- 111
- Issue
- 3
- Pages
- 257-276
- ISSN
- 0018-9219 (ISSN-L)
- 1558-2256
- OpenAlex
- W4320002812
- Cited by
- 3,689
- References
- 272
- Citations by year
- 2026: 610 · 2025: 1013 · 2024: 940 · 2023: 541
- FWCI
- 280.65
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- Proceedings of the IEEE (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Natural Science Foundation of China (62125102)
- National Key Research and Development Program of China
- Fundamental Research Funds for the Central Universities
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Visual Attention and Saliency Detection
- Keywords
- object detection, evaluation metrics, deep learning
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.