What Makes for Effective Detection Proposals?
Open accessThe DOI matches its Crossref record; 69 of the 69 references it cites were found in a registry or an index.
Abstract
Current top performing object detectors employ detection proposals to guide the search for objects, thereby avoiding exhaustive sliding window search across images. Despite the popularity and widespread use of detection proposals, it is unclear which trade-offs are made when using them during object detection. We provide an in-depth analysis of twelve proposal methods along with four baselines regarding proposal repeatability, ground truth annotation recall on PASCAL, ImageNet, and MS COCO, and their impact on DPM, R-CNN, and Fast R-CNN detection performance. Our analysis shows that for object detection improving proposal localisation accuracy is as important as improving recall. We introduce a novel metric, the average recall (AR), which rewards both high recall and good localisation and correlates surprisingly well with detection performance. Our findings show common strengths and weaknesses of existing methods, and provide insights and metrics for selecting and tuning proposal methods.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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
“It is noteworthy that the Recall-to-IoU metric is just loosely [19], [20], [21] related to the ultimate detection accuracy.”
From Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (2017) · cites this paper as [20]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In the case of sliding window based detectors, the imbalance between backgrounds and objects could be as extreme as 10 7 :1 [71].”
From Object Detection in 20 Years: A Survey (2023) · cites this paper as [71]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
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 1×
“Indeed, there has been some very recent work in this space [60], [61], [62] that shows promising”
BackgroundFaster 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 · cited 1×
“Indeed, there has been some very recent work in this space [60], [61], [62] that shows promising”
BackgroundImageNet: 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
Microsoft COCO: Common Objects in Context
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10602-1_48 · Cited by 45,460
Self-citationRich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
Girshick, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.169 · Cited by 28,105
Robust Real-Time Face Detection
Viola, P., & Jones, M. J. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000013087.49260.fb · Cited by 14,097
Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2015.2389824 · Cited by 11,971
Object Detection with Discriminatively Trained Part-Based Models
Felzenszwalb, P. F., Girshick, R. B., McAllester, D., & Ramanan, D. (2010) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2009.167 · Cited by 9,853
The Pascal Visual Object Classes Challenge: A Retrospective
Everingham, M., Eslami, S. M. A., Van Gool, L., Williams, C. K. I., Winn, J., & Zisserman, A. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-014-0733-5 · Cited by 7,139 · cited 3×
“We test the generalisation of proposal methods by evaluating on datasets with many additional classes in §4.”
Method
Research works printed but not linked: 7
- [28] A. Humayun, F. Li, and J. M. Rehg, 'Rigor: Recycling inference in graph cuts for generating object regions,' in CVPR , 2014. -
- [22] P. Krähenbühl and V. Koltun, 'Geodesic object proposals,' in ECCV , 2014. -
- [51] P. Dollár, R. Appel, and W. Kienzle, 'Crosstalk cascades for framerate pedestrian detection,' in ECCV , 2012. -
- [59] K. Tang, A. Joulin, L.-J. Li, and L. Fei-Fei, 'Co-localization in realworld images,' in CVPR , 2014. -
- [57] S. Vicente, C. Rother, and V. Kolmogorov, 'Object cosegmentation,' in CVPR , 2011. -
- [54] N. Chavali, H. Agrawal, A. Mahendru, and D. Batra, 'Objectproposal evaluation protocol is 'gameable',' arXiv:1505.05836 , 2015. - · arXiv
- szegedy, 0, Scalable, high-quality object detection, arXiv 1412 1441
Cited by
Items 501–600 of 764
Real-time Human Detection based on Personness Estimation
Kim, K., & Sohn, K. (2015) · British Machine Vision Conference (BMVC) · doi:10.5244/c.29.113 · Cited by 5
“detection proposals) measure methods have gained popularity recently [10].”
MethodInfluentialAlgorithm for diabetic retinal image analysis based on deep learning
Deng, L., Liu, S.-S., Cheng, Y., Zhao, G., & Xu, J. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-15503-w · Cited by 4
“…between the anchor i and the ground-truth box is >0.7, and p ∗ i = 1 ; if the anchor is negative, then the IoU between the anchor i and the ground-truth box is <0.3, where p ∗ i = 0 . t is the predicted bounding box’s vector, and t * is the vector corresponding to the ground-truth box [18, 19].”
BackgroundInfluentialAction Progression Networks for Temporal Action Detection in Videos
Lu, C., Mak, M., Li, R., Chi, Z., & Fu, H. C. (2024) · IEEE Access · doi:10.1109/access.2024.3451503 · Cited by 5
“More-over, it presents a notable trade-off between computational requirements and high-quality detection [20], [21].”
BackgroundDetection and Identification of Hazardous Hidden Objects in Images: A Comprehensive Review
Swain, S., & Devi, K. S. (2024) · Archives of Computational Methods in Engineering · doi:10.1007/s11831-024-10173-9 · Cited by 7
“The disparity between backgrounds and objects with respect to sliding window-based detectors might be as high as 10 7 :1 [78].”
Suganob, N. J., Concepcion, R., Billones, R. K. C., & Dadios, E. P. (2024) · Lecture notes in networks and systems · doi:10.1007/978-3-031-73321-5_11 · Cited by 4
Food Recognition and Detection with Minimum Supervision
Liu, R. (2016) · Scholarship@Western (Western University) · Cited by 5
Huang, S., Wang, W., He, S., & Lau, R. W. H. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2627819 · Cited by 4
“(See [17] for a comprehensive review.)”
Personness estimation for real-time human detection on mobile devices
Kim, K., Oh, C., & Sohn, K. (2016) · Expert Systems with Applications · doi:10.1016/j.eswa.2016.12.017 · Cited by 4
“To examine the feasibility of the existing methods on obile devices, we measured the running times of the following asic preprocessing algorithms on a mobile device: 1) Structured dges ( SE ) ( Dollár & Zitnick, 2013 ): a state-of-the-art algorithm or edge detection adopted by RIGOR ( Humayun et al., 2014 ), OP ( Kräh…”
BackgroundMethodInfluentialKey Indicators to Assess the Performance of LiDAR-Based Perception Algorithms: A Literature Review
Chiranjeevi, K., da Silva, J. M., & Correia, M. V. (2023) · IEEE Access · doi:10.1109/access.2023.3321912 · Cited by 4
“The aggressiveness of object detectors for a particular class is measured using another assessment metric called average recall [50].”
BackgroundCorner points localization in electronic topographic maps with deep neural networks
Dong, L., Zheng, F., Chang, H., & Qin, Y. (2017) · Earth Science Informatics · doi:10.1007/s12145-017-0317-3 · Cited by 4
“One is the object proposal (Gidaris and Komodakis, 2016c; Hosang et al., 2016; Kong et al., 2016; Krähenbühl and Koltun, 2014, 2015; Ren et al., 2015).”
BackgroundCascaded convolutional neural networks for object detection
Guo, Y., Guo, X., Jiang, Z., & Zhou, Y. (2017) · IEEE International Conference on Visual Communications and Image Processing (VCIP) · doi:10.1109/vcip.2017.8305026 · Cited by 4
“[8] provide an in-depth analysis of twelve proposal methods along with four baselines, showing that Selective Search [5] and EdgeBoxes [6] are effective methods which compromise speed, accuracy, and simplicity.”
MethodDEEP CROSS-DOMAIN BUILDING EXTRACTION FOR SELECTIVE DEPTH ESTIMATION FROM OBLIQUE AERIAL IMAGERY
Ruf, B., Thiel, L., & Weinmann, M. (2018) · ISPRS annals of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprs-annals-iv-1-125-2018 · Cited by 4
“Thereby, the grouping is typically based on a diverse set of cues including superpixel shape, appearance cues and boundary estimates (Hosang et al., 2016; Sommer et al., 2016).”
BackgroundSQL: Superpixels via quaternary labeling
Chai, D. (2019) · Pattern Recognition · doi:10.1016/j.patcog.2019.03.012 · Cited by 4
Research on Small Target Recognition Technology Based on Deep Learning
Li, Y., Jian-wang, H., & Ji, B. (2019) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1237/2/022119 · Cited by 4
Traffic Lights Detection Based on Deep Learning Feature
Wang, C., Zhang, G., Zhou, W., Rao, Y., & Lv, Y. (2020) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-030-44751-9_32 · Cited by 4
Efficient Proposals: Scale Estimation for Object Proposals in Pedestrian Detection Tasks
Qiu, J. W., Wang, L., Wang, Y., & Hu, Y. H. (2020) · IEEE Signal Processing Letters · doi:10.1109/lsp.2020.2995102 · Cited by 4
“A comprehensive survey and comparison of object proposal generation methods can be found in [20].”
BackgroundSmall target detection based on bird’s visual information processing mechanism
Wang, Z., Liu, D., Lei, Y., Niu, X., Wang, S., & Shi, L. (2020) · Multimedia Tools and Applications · doi:10.1007/s11042-020-08807-8 · Cited by 4
A class-independent flexible algorithm to generate region proposals
Taghizadeh, M., & Chalechale, A. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-10826-y · Cited by 4
“Moreover, they indicted much lower AR in experiments [17].”
BackgroundPneumonia Detection using Ensemble of Modified ViT-YOLO Models
Chetwani, S. A., Vaidya, K. P., & Dhabu, M. M. (2022) · International Conference on Computing, Communication and Networking Technologies (ICCCNT) · doi:10.1109/icccnt54827.2022.9984598 · Cited by 4
Metrics for Machine Learning Models to Facilitate SOTIF Analysis in Autonomous Vehicles
Madala, K., & Gonzalez, C. A. (2023) · SAE International Journal of Advances and Current Practices in Mobility · doi:10.4271/2023-01-0829 · Cited by 4
Rapid Fog-Removal Strategies for Traffic Environments
Liu, X., Hong, L., & Lin, Y. (2023) · Sensors · doi:10.3390/s23177506 · Cited by 4
“The representative algorithms are R-CNN [23], Mask R-CNN [24], Fast R-CNN [25], Faster R-CNN [26], etc.”
MethodSemantic segmentation priors for object discovery
García, G. M., Husain, F., Schulz, H., Frintrop, S., Torras, C., & Behnke, S. (2016) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2016.7899691 · Cited by 4
“[6] show that one of the most successful methods in terms of recall is the Selective Search approach of Uijlings et al.”
MethodTarget recognition of ground armor based on combinatorial part model
Chen, S., Ren, M., & Chen, D. (2017) · IEEE International Conference on Communication Software and Networks · doi:10.1109/iccsn.2017.8230301 · Cited by 4
Fast deep pyramid DPM object detection with region proposal networks
Ali, A., Olaleye, O. G., Dey, B., & Bayoumi, M. (2017) · IEEE International Symposium on Signal Processing and Information Technology (ISSPIT) · doi:10.1109/isspit.2017.8388636 · Cited by 4
“Recent research presents different region proposal techniques [18] including the state-of-the-art selective search, objectness, Binarized normed gradients (BING), constrained parametric min-cuts (CPMC) and Edge Boxes [18] to generate elementary candidates and reduce the anticipated search space in object detection pro…”
A two-stage urine sediment detection method
Wang, Q., Sun, Q., & Wang, Y. (2020) · International Conference on Image, Video Processing and Artificial Intelligence (IVPAI) · doi:10.1117/12.2577493 · Cited by 4
“The candidate region proposal algorithm mainly performs classifications in two ways, namely the combination method and the window scoring method.(15) This paper mainly discusses the former method which aims to break and aggregate the image, including Selective Search and Edge Boxes.”
MethodCandidate region acquisition optimization algorithm based on multi-granularity data enhancement
Chen, D., Miao, D., Zhao, C., & Zhou, H. (2022) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-021-01492-5 · Cited by 4
“SS is still a relatively common and reliable method to obtain RP [10], but there are some shortcomings of the SS method that need to be compensated in terms of the image data used.”
BackgroundMethodInfluentialAn improved attention mechanism based YOLOv4 structure for lung nodule detection
Wu, D., Lü, T., & Li, X. (2022) · 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/cy55231.2022.9927936 · Cited by 4
“Calculate the accuracy detection metric AP (Average Precision) [28] to evaluate the model.”
MethodRegion Trajectories for Video Semantic Concept Detection
Ye, Y., Rong, X., Yang, X., & Tian, Y. (2016) · International Conference on Multimedia Retrieval · doi:10.1145/2911996.2912046 · Cited by 3
“Selective Search [22] and EdgeBox [28] are combined in our implementation to generate region proposals [7], which can improve the coverage rate of the objects in the image.”
Object Proposal via Depth Connectivity Constrained Grouping
Wang, Y., Huang, L., Ren, T., Zhong, S.-H., Liu, Y., & Wu, G. (2018) · Lecture notes in computer science · doi:10.1007/978-3-319-77383-4_4 · Cited by 3
Region Proposal Ranking via Fusion Feature for Object Detection
Li, X., Ma, H., & Wang, X. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451326 · Cited by 3
“Region proposal is an objectness area extraction method based on simulating human attention mechanisms, and it is widely used in the field of computer vision, especially object detection tasks [1], [2], [3], [4].”
MethodLi, C., Huang, H., Mo, H., Wen, Z., Zhou, S., & Zhu, Z. (2025) · IEEE Access · doi:10.1109/access.2025.3581313 · Cited by 5
“This network extracts object candidate boxes [14] through selective search, resizes each proposal, and uses a CNN model to extract features.”
Generating object proposals for improved object detection in aerial images
Sommer, L. W., Schuchert, T., & Beyerer, J. (2016) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2241527 · Cited by 3
“In recent years, object proposals methods have been proposed as an alternative approach to reduce the number of candidate windows compared to sliding window approaches [1].”
MethodA Real-time Pedestrian Detection System in Street Scene
Guo, A., Xu, M., Ran, F., & Wang, Q. (2016) · International Journal on Smart Sensing and Intelligent Systems · doi:10.21307/ijssis-2017-931 · Cited by 3
“The goal of ADAS is to achieve zero fatalities through an active or passive safety system[1][2].”
BackgroundAWBING plus algorithm for generic object proposal generation
Liu, Q., Yang, F., & Li, C. (2019) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-18810 · Cited by 3
“Such runtime efficiency as well as its high DR, remarkably surpass other well-known methods [24].”
Saliency detection for improving object proposals
Chen, S., Li, J., Hu, X., & Zhou, P. (2016) · IEEE International Conference on Digital Signal Processing (DSP) · doi:10.1109/icdsp.2016.7868600 · Cited by 3
Object proposal using 3D point cloud for DRC-HUBO+
Shin, S., Shim, I., Jung, J., Bok, Y., Oh, J.-H., & Kweon, I. S. (2016) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2016.7759113 · Cited by 3
“Selective Search [11] and Edge Boxes [21] show consistently good performance both in ground truth recall and high repeatability in the evaluation of [6].”
BackgroundMethodRich Features and Precise Localization with Region Proposal Network for Object Detection
Chu, M., Wu, S., Gu, Y., & Xu, Y. D. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-69923-3_65 · Cited by 3
“Object detection is one of the most fundamental researches in computer vision [6][7].”
Common visual pattern discovery and search
Wang, Z., Meng, J., Yu, T., & Yuan, J. (2017) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipa.2017.8282178 · Cited by 3
“, the object proposals [40], [39].”
An Obstacle Detection Method Based on Binocular Stereovision
Sun, Y., Zhang, L., Leng, J., Luo, T., & Wu, Y. (2018) · Lecture notes in computer science · doi:10.1007/978-3-319-77383-4_56 · Cited by 3
A Minimum Barrier Distance based Saliency Box for Object Proposals Generation
Huang, X., Zheng, Y., Huang, J., & Zhang, Y. (2018) · IEEE Signal Processing Letters · doi:10.1109/lsp.2018.2844097 · Cited by 3
“AR reflects the average object proposals generation performance across different IoU thresholds for each model and [19] shows that AR correlates well with the detection performance.”
BackgroundInfluentialMulti ROI and Multi Map Networks for Accurate and Efficient Pedestrian Detection
Qiu, Z., & Gu, X. (2018) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2018.8489310 · Cited by 3
“Comparisons and evaluations can be found in [4-6].”
BackgroundDetecting Buildings of any Size Using Integration of Cnn Models
Hamaguchi, R., Nemoto, K., Imaizumi, T., & Hikosaka, S. (2018) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss.2018.8517887 · Cited by 3
Multiobject Detection Algorithm Based on Adaptive Default Box Mechanism
Li, J., Hou, Q., & Xing, J. (2020) · Complexity · doi:10.1155/2020/5763476 · Cited by 4
“Taking FPS, mAP, andmAR [36] as evaluation criteria, the experimental comparison results of the six models are shown in Table 2.”
MethodRetractedImage Segmentation and Object Recognition
Kumar, A. S. (2021) · Cited by 3
A Novel Electronic Chip Detection Method Using Deep Neural Networks
Zhang, H., Sun, H., Shi, P. J., & Minchala, L. I. (2022) · Machines · doi:10.3390/machines10050361 · Cited by 3
A comparative analysis of optical sensor artifacts across neural network-based ATR algorithms
Davidson, Q. T. (2024) · Automatic target recognition. Conference · doi:10.1117/12.3014530 · Cited by 3
Deep learning-based modified-EAST scene text detector: insights from a novel multiscript dataset
Mahajan, S., Rani, R., & Kamboj, A. (2024) · International Journal on Document Analysis and Recognition (IJDAR) · doi:10.1007/s10032-024-00491-w · Cited by 3
“Additionally, the significance of IoU in evaluating the effectiveness of the object detection model has been covered by many experts [14, 18, 53].”
APPLICABILITY EVALUATION OF OBJECT DETECTION METHOD TO SATELLITE AND AERIAL IMAGERIES
Kamiya, K., Fuse, T., & Takahashi, M. (2016) · The international archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences · doi:10.5194/isprsarchives-xli-b7-229-2016 · Cited by 3
Pei, T., & Wu, X. (2016) · IEEE International Conference on Computer and Communications (ICCC) · doi:10.1109/compcomm.2016.7924740 · Cited by 3
Lin, G., & Qin, S. (2017) · International Conference on Control and Automation · doi:10.1109/icca.2017.8003209 · Cited by 3
A Real-Time Traffic Light Detection Algorithm Based on Adaptive Edge Information
Yu, G., Ao, L., Li, H., Wang, Y., Wang, Z., & Hu, C. (2018) · SAE technical papers on CD-ROM/SAE technical paper series · doi:10.4271/2018-01-1620 · Cited by 3
Parallel Search by Reinforcement Learning for Object Detection
Huang, Y., Gu, C., Wu, K., & Guan, X. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-03341-5_23 · Cited by 3
Siamese Tracking from Single Point Initialization
Zheng, X., Luo, H., Hui, B., & Chang, Z. (2019) · Sensors · doi:10.3390/s19030514 · Cited by 3
Target Detection and Recognition Based on Improved Faster R-CNN
房, 靖. (2019) · Journal of Image and Signal Processing · doi:10.12677/jisp.2019.82007 · Cited by 3
Cross-layer fusion feature network for material defect detection
Yang, K., Sun, Z., Wang, A., Liu, R., Liu, L., & Wang, Y. (2019) · Journal of Electronic Imaging · doi:10.1117/1.jei.28.3.033025 · Cited by 3
“In the region proposal algorithm, region proposals are generated from a CNN to realize end-to-end training in MultiBox with good performance.24 DeepBox is trained with a ConvNet model that learns to rerank region proposals generated by Edge Boxes.25,26 For object detection based on the regression method, Shen et al.27…”
BackgroundFast Efficient Object Detection Using Selective AttentionPreprint
Yohanandan, S. A. C., Song, A., Dyer, A. G., Faragasso, A., Roy, S. S., & Tao, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.07502 · Cited by 3
“As pioneered in the Selective Search work [24], the first stage generates a sparse set of ideally object-only candidate proposals while filtering out the majority of negative locations [25], while the second stage classifies the proposals into object-category classes.”
BackgroundGeneric instance segmentation for object-oriented bin-picking
Grard, M. (2019) · theses.fr (ABES) · doi:10.70675/60637975z66b0z4c66za077zbda1a1636ab7 · Cited by 8
Sampaio, A. F. (2019) · Portuguese National Funding Agency for Science, Research and Technology (RCAAP Project by FCT) · doi:10.34626/3j6w-a440 · Cited by 3
Object proposal with kernelized partial ranking
Wang, J., Shen, J., & Li, P. (2017) · Pattern Recognition · doi:10.1016/j.patcog.2017.03.022 · Cited by 2
“sses and has standard subset for training, validation and testing. We use the first two subsets for our training. Evaluation Metrics. The employed evaluation metrics are Recall and Average Recall (AR) [HBDS16]. We say that a proposal contains an object if it is above a given intersection over union (IoU) threshold. Ave…”
MethodResultInfluentialDynamic Routing and Calibration for Few-Shot Object Detection
Wu, J., Lei, J., Tian, H., Liu, X., Feng, Z., & Liang, R. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10888850 · Cited by 2
Maviane-Macia, F., Wiedemann‐Merdinoglu, S., Rousseau, D., & Peeters, N. M. (2025) · Plant Methods · doi:10.1186/s13007-025-01432-2 · Cited by 2
Motion estimation for multi-object tracking using KalmanNet with semantic-independent encoding
Song, J., Mei, W., Xu, Y., Fu, Q., Kou, R., Bu, L., & Long, Y. (2026) · Information Fusion · doi:10.1016/j.inffus.2026.104513 · Cited by 3
“5 , 1] , following the convention in object detection [39].”
Research on an industrial liquid level detection method based on an improved SSD
Qi, X., & Jia, L. (2025) · Scientific Reports · doi:10.1038/s41598-025-29277-8 · Cited by 2
Wilson, K. C., Lurbur, M., & Yochum, N. (2025) · ICES Journal of Marine Science · doi:10.1093/icesjms/fsaf168 · Cited by 2
Deep Learning-Driven Segmentation of Unseen Objects in Indoor Robotic Environments
Zhang, Y., Yin, M., Zhang, Q., Zhang, C.-H., Yang, S. X., Li, Z., & Hua, C. (2026) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2026.3673587 · Cited by 4
“Average Recall (AR): AR evaluates the quality of region proposals by measuring the maximum recall for a fixed number of detections per image, averaged over the same IoU thresholds as AP [104], [105].”
Image foreground extraction — A supervised framework based on region transfer
Banerjee, B., Saha, S., & Merchant, S. N. (2016) · International Conference on Signal and Information Processing · doi:10.1109/iconsip.2016.7857485 · Cited by 2
“Region proposal generation techniques [4] [10] [11] [12] [13] are employed for this purpose to highlight some specific locations in the image where the probability of having an object instance is high.”
MethodImproving object proposals with top-down cues
Li, W., Li, H., Luo, B., Shi, H., Wu, Q., & Ngan, K. N. (2017) · Signal Processing Image Communication · doi:10.1016/j.image.2017.04.006 · Cited by 2
Improved Traffic Sign Detection In Videos Through Reasoning Effective RoI Proposals
Zhang, Y., Qi, Y., Yang, J., & Hwang, J. (2020) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme46284.2020.9102896 · Cited by 2
“In general, region-based methods, which pass the proposal candidates with a higher object-containing probability to the subsequent object clas-sification and bounding box regression tasks, can produce better detection performance compared with the single-stage methods [11].”
ResultWeakly Supervised Sentiment-Specific Region Discovery for VSA
Xue, L., Xu, A., Mao, Q., Gao, L., & Chen, J. (2020) · The Computer Journal · doi:10.1093/comjnl/bxaa112 · Cited by 2
Device Target Checking for Power Patrol Robot Based on Objectness Estimation
Huang, L., Ma, X., Fang, F., & Zhou, B. (2020) · IEEE Conference on Industrial Electronics and Applications (ICIEA) · doi:10.1109/iciea48937.2020.9248095 · Cited by 2
“After filtering proposals with non-maximal suppression (NMS) [13], the final objectess score for each size of windows can be obtained from (3): where , i v In order to obtain a window closer to the size of the object in the image, the proposals need to be further filtered.”
MethodInsulator Fault Recognizing via Modified Faster R-CNN Using UAV Data
Chen, Y., Yang, X., Liu, H., Gui, Y., Li, W., & Qiu, Q. (2021) · International Conference on Wireless Communications and Smart Grid (ICWCSG) · doi:10.1109/icwcsg53609.2021.00023 · Cited by 2
A GAN-based small target detection model
Fu, J., & Chen, Z. (2023) · International Conference on Computer Vision and Data Mining (ICCVDM) · doi:10.1117/12.2660096 · Cited by 2
“Directly using this method to generate proposals will have larger problems for small object detection [24].”
MethodGated Enhanced RPN and Hybrid-View for Few-Shot Object Detection
Wei, X., Zhou, Z., Guo, P., & Zhang, W. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096737 · Cited by 2
“The visualization comparison between the proposed method and the latest methods [10, 13] for 10-shot case. first stage is the Region Proposal Network (RPN) [12].”
MethodMultipath Cluster-Based Scatterer Recognition by Object Detection Techniques Using Panoramic Images
Calist, I., & Kim, M. (2023) · International Technical Conference on Circuits/Systems, Computers and Communications (ITC-CSCC) · doi:10.1109/itc-cscc58803.2023.10212703 · Cited by 2
Jacob, B., McDonald, H., & Bohn, J. (2024) · American Journal of Neural Networks and Applications · doi:10.11648/j.ajnna.20241001.11 · Cited by 2
Feasibility of the Anchor‐Free Deep Learning Method in Coronary Stenosis Automatic Detection
Yue, H., Yu, W., Dong, J., Lai, Y., Wu, Y., Zhao, H., Song, Y., Zhao, L., Wang, H., Zhang, J., et al. (2024) · Journal of Interventional Cardiology · doi:10.1155/2024/2606789 · Cited by 2
Aiming Error Analysis of Controlled Turret System With AI Target Recognition
Karlson, M., Ban, H., Cole, D. G., Abdelhakim, M., & Forsythe, J. (2025) · International Journal of Control Automation and Systems · doi:10.1007/s12555-024-1015-9 · Cited by 2
“5 , 0 ) [38], where G is the number of ground truth objects in D .”
Improving an object tracker for infrared flying bird tracking
Huang, Y. P., Zheng, H. Y., & Yang, H. (2016) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2016.7532648 · Cited by 2
Object proposals using CNN-based edge filtering
Waris, M. A., Iosifidis, A., & Gabbouj, M. (2016) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2016.7899704 · Cited by 2
“Following [7], [34], we also evaluate the quality of object proposals using the recall metric.”
BackgroundMethodMulti-scale energy optimization for object proposal generation
Wang, C., Yang, J., Wang, K., & Lai, S. (2016) · Multimedia Tools and Applications · doi:10.1007/s11042-016-3616-7 · Cited by 2
Category independent object proposals using quantum superposition
Malik, J., Aytekin, Ç., & Gabbouj, M. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8297068 · Cited by 2
An improved real-time object proposals generation method based on local binary pattern
Jiang, Y., Yan, J., Fan, C., Shi, W., & Deng, D. (2017) · International Journal of Advanced Robotic Systems · doi:10.1177/1729881417724679 · Cited by 2
The Optimal pedestrian detection algorithm based on dynamic adaptive region convolution model
Qiu, D., & Liu, D. (2017) · Chinese Automation Congress · doi:10.1109/cac.2017.8244213 · Cited by 2
A hierarchical inferential method for indoor scene classification
Jiang, J., Liu, P., Ye, Z., Zhao, W., & Tang, X. (2017) · International Journal of Applied Mathematics and Computer Science · doi:10.1515/amcs-2017-0059 · Cited by 2
“…representation (Rigamonti et al., 2011; 2013), neural networks (Liu et al., 2016; Kong et al., 2016; Bell et al., 2016; He et al., 2016) and the SVM (Felzenszwalb et al., 2010b; 2008; 2010a; Sadovnik and Chen, 2011; Felzenszwalb and McAllester, 2007; Girshick et al., 2011; Hosang et al., 2016).”
Figure-ground segmentation based on class-independent shape priors
Li, Y., Liu, Y., Liu, G., & Guo, M. (2018) · Journal of Electronic Imaging · doi:10.1117/1.jei.27.1.013018 · Cited by 2
“The average recall simultaneously measures proposal recall value and location accuracy.(52) The MCG scheme learns a grouping strategy based on pixel-level annotations; POISE develops various seeds to segment objects, and although it has no training process, segmentation performance is affected by various parameter set…”
MethodVisual saliency using unit-linking PCNN image segmentation
Hua, Y., & Gu, X. (2017) · International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD) · doi:10.1109/fskd.2017.8393392 · Cited by 2
An Effective Zoom-In Approach for Detecting DIM and Small Target Proposals in Satellite Imagery
Zhang, J., Jia, X., & Hu, J. (2018) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss.2018.8519575 · Cited by 2
“A baseline experiment was conducted by extracting target proposals directly from segmentation [5].”
MethodFully Convolutional Network With Densely Feature Fusion Models for Object Detection
Huang, S., Liy, X., Jiang, Z., Guo, X., & Men, A. (2018) · IEEE International Conference on Multimedia & Expo Workshops (ICMEW) · doi:10.1109/icmew.2018.8551503 · Cited by 2
“Proposals are used to guide the search for possible object locations, and a variety of papers offer methods for generating category-independent region proposals [19].”
Lu, Q., Jiang, Z., Men, A., & Tang, P. (2019) · IET Computer Vision · doi:10.1049/iet-cvi.2018.5601 · Cited by 2
Accelerate proposal generation in R-CNN methods for fast pedestrian extraction
Wang, J., & Li, G. (2019) · The Electronic Library · doi:10.1108/el-09-2018-0191 · Cited by 2
Structured Light Detection Algorithm based on Deep Learning
Ma, S., Song, Y., Cheng, N., Hao, Y., Chen, Z., & Fu, X. (2019) · IOP Conference Series Earth and Environmental Science · doi:10.1088/1755-1315/252/4/042050 · Cited by 2
Sugarcane Node Identification Based on Structured Learning Model
Hu, X., Li, S., Wang, M., Shi, C., Shang, J., Pei, Y., Deng, G., & Zhou, D. (2019) · Communications in computer and information science · doi:10.1007/978-981-13-9917-6_13 · Cited by 2
Zhang, Y., Hilton, A., & Guillemaut, J. (2020) · International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP) · doi:10.5220/0008991104520461 · Cited by 2
“We refer the reader to recent surveys for a more comprehensive review (Zhao et al., 2018; Li et al., 2015; Hosang et al., 2016; Neelima et al., 2015).”
Infrared Image Fault Identification of Power Equipment Based on Residual Network
Zhou, F., Ma, Y., Ma, Y., & Pan, H. (2020) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-44038-1_1 · Cited by 2
Vehicle Detection Using Faster Recurrent Convolution Neural Network
Chate, M., & Gohokar, V. V. (2020) · Lecture notes in electrical engineering · doi:10.1007/978-981-15-1420-3_130 · Cited by 2
A Study of Dimensionality Reduction Impact on an Approach to People Detection in Gigapixel Images
Ferreira, C. B. R., Soares, F. A. A. D. M. N., Pedrini, H., Bruce, N. D. B., Ferreira, W. D., & da Cruz, G. (2020) · Canadian Journal of Electrical and Computer Engineering · doi:10.1109/cjece.2019.2925780 · Cited by 2
“Thus, in order to reduce unnecessary bounding boxes, we perform a merging process by calculating the intersection over union (IoU) based on a nonmaximum suppression criterion [18], [19].”
MethodRETRACTED ARTICLE: UAV landmark detection on fast region-based CNN
Yang, R. (2021) · Arabian Journal of Geosciences · doi:10.1007/s12517-021-07457-w · Cited by 2
RetractedCargo Vehicle Classification System Through Axle Detection
Vazquez, M., Cikel, K., Arzamendia, M., Gregor, D. O., & Villagra, M. (2020) · IEEE Congreso Bienal de Argentina (ARGENCON) · doi:10.1109/argencon49523.2020.9505351 · Cited by 2
“This information is included in the output for each detected object in the images [14].”
BackgroundA deep learning-based and adaptive region proposal algorithm for semantic segmentation
Taghizadeh, M., & Chalechale, A. (2024) · Applied Soft Computing · doi:10.1016/j.asoc.2024.111410 · Cited by 3
The Image Torque Operator for Contour ProcessingPreprint
Nishigaki, M., & Fermuller, C. M. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1601.04669 · Cited by 2
Metadata
- Authors
- Hosang, J., Benenson, R., Dollar, P., & Schiele, B.
- Affiliations
- Max Planck Institute for Informatics (DE)
- ORCID
- Schiele, Bernt
- Venue
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Published
- 2016-04-01
- Volume
- 38
- Issue
- 4
- Pages
- 814-830
- ISSN
- 0162-8828 (ISSN-L)
- 2160-9292
- OpenAlex
- W1958328135
- Cited by
- 727
- References
- 68
- Citations by year
- 2026: 9 · 2025: 27 · 2024: 34 · 2023: 36 · 2022: 57 · 2021: 58 · 2020: 63 · 2019: 95 · 2018: 93 · 2017: 126 · 2016: 98
- FWCI
- 100.45
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Transactions on Pattern Analysis and Machine Intelligence (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- MPG.PuRe (Max Planck Society) (submittedVersion, other-oa)
- MPG.PuRe (Max Planck Society) (submittedVersion)
- arXiv (Cornell University)
- 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, PASCAL dataset, ImageNet, MS COCO, DPM, R-CNN, Fast R-CNN, sliding window search
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.
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.