Multi-cue onboard pedestrian detection
The DOI matches its Crossref record; 36 of the 37 references it cites were found in a registry or an index.
Abstract
Various powerful people detection methods exist. Surprisingly, most approaches rely on static image features only despite the obvious potential of motion information for people detection. This paper systematically evaluates different features and classifiers in a sliding-window framework. First, our experiments indicate that incorporating motion information improves detection performance significantly. Second, the combination of multiple and complementary feature types can also help improve performance. And third, the choice of the classifier-feature combination and several implementation details are crucial to reach best performance. In contrast to many recent papers experimental results are reported for four different datasets rather than using a single one. Three of them are taken from the literature allowing for direct comparison. The fourth dataset is newly recorded using an onboard camera driving through urban environment. Consequently this dataset is more realistic and more challenging than any currently available dataset.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
Histograms of Oriented Gradients for Human Detection
Dalal, N., & Triggs, B. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.177 · Cited by 33,584 · cited 5×
“Dalal et al. [6] enrich their static feature descriptor [ 5 ] with internal motion histograms to improve detection.”
BackgroundMethodResultInfluentialThe Ecological Approach to Visual Perception
Bornstein, M. H., & Gibson, J. J. (1980) · The Journal of Aesthetics and Art Criticism · doi:10.2307/429816 · Cited by 22,935 · cited 1×
“While psychologists and neuroscientists argue that motion is an important cue for human perception [17] only few computer vision object detectors (e.g. [30, 6]) exploit this fact.”
BackgroundRobust 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
Friedman, J., Hastie, T., & Tibshirani, R. (2000) · The Annals of Statistics · doi:10.1214/aos/1016218223 · Cited by 6,884 · cited 1×
“AdaBoost Contrary to SVMs, boosting algorithms [14] optimize the classification error on the training samples iteratively.”
MethodA discriminatively trained, multiscale, deformable part model
Felzenszwalb, P., McAllester, D., & Ramanan, D. (2008) · 2008 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2008.4587597 · Cited by 2,867 · cited 2×
“Popular classifiers are SVMs [26, 5, 13, 20, 19] or boosting [31, 7, 33, 23].”
BackgroundDetecting Pedestrians Using Patterns of Motion and Appearance
Viola, P., Jones, M. J., & Snow, D. (2005) · International Journal of Computer Vision · doi:10.1007/s11263-005-6644-8 · Cited by 2,079
A Duality Based Approach for Realtime TV-L 1 Optical Flow
Zach, C., Pock, T., & Bischof, H. (2007) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-74936-3_22 · Cited by 1,658 · cited 2×
“de Objectives Detect pedestrians from a moving platform • Exploit motion information • Leverage complementarity of features • Evaluate different classifiers • New datasets with image pairs Features • HOG [1] 8× 8 pixel cells, 2× 2 blocks 9-bin histograms, unsigned gradients • Haar wavelets [2] 32 and 16 pixel masks ho…”
BackgroundHuman Detection Using Oriented Histograms of Flow and Appearance
Dalal, N., Triggs, B., & Schmid, C. (2006) · Lecture Notes in Computer Science · doi:10.1007/11744047_33 · Cited by 1,652 · cited 7×
“In contrast, [ 6 ] proposed motion features that are ‐ at least in principle ‐ applicable to onboard sequences.”
BackgroundMethodResultInfluentialPedestrian detection: A benchmark
Dollar, P., Wojek, C., Schiele, B., & Perona, P. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206631 · Cited by 1,313 · cited 3×
“Doll´ ar et al. [ 8 ] report results for various recent approaches on a new challenging onboard dataset with improved evaluation metrics.”
BackgroundInfluentialSelf-citationA Trainable System for Object Detection
Papageorgiou, C., & Poggio, T. (2000) · International Journal of Computer Vision · doi:10.1023/a:1008162616689 · Cited by 1,289 · cited 3×
“Haar Haar wavelets have been introduced by Papageor-giou&Poggio [22] for people detection.”
BackgroundMethodInfluential
Research works printed but not linked: 1
- maji, 2008, Classification using intersection kernel SVMs is efficient, CVPR
Cited by
Items 101–200 of 370
Review of pedestrian detection techniques in automotive far‐infrared video
Hurney, P., Waldron, P., Morgan, F., Jones, E., & Glavin, M. (2015) · IET Intelligent Transport Systems · doi:10.1049/iet-its.2014.0236 · Cited by 51
Learning a Part-Based Pedestrian Detector in a Virtual World
Xu, J., Vázquez, D., López, A. M., Marín, J., & Ponsa, D. (2014) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2014.2310138 · Cited by 58
“Since our interest is pedestrian detection for cars, we validate our proposals in different data sets acquired on board, namely Daimler [1], TUD-Brussels [29], Caltech-Testing [3], and CVC02 [30].”
MethodExpecting the Unexpected: Training Detectors for Unusual Pedestrians with Adversarial Imposters
Huang, S., & Ramanan, D. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.496 · Cited by 49
“Pedestrian detection is a highly researched topic with mature methods, but most datasets focus on “everyday” scenes of people engaged in typical walking poses on sidewalks [9, 6, 13, 12, 48].”
BackgroundMachine learning-based augmented reality for improved surgical scene understanding
Pauly, O., Diotte, B., Fallavollita, P., Weidert, S., Euler, E., & Navab, N. (2014) · Computerized Medical Imaging and Graphics · doi:10.1016/j.compmedimag.2014.06.007 · Cited by 48
Automatic detection and tracking of pedestrians from a moving stereo rig
Schindler, K., Ess, A., Leibe, B., & Gool, L. V. (2010) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2010.06.006 · Cited by 53
“In future work, we plan to also include optic266 flow between consecutive frames, which has been shown to consistently improve267 detection in a dynamic environment (Wojek et al., 2009; Walk et al., 2010).”
MethodEfficient Stixel-based object recognition
Enzweiler, M., Hummel, M., Pfeiffer, D., & Franke, U. (2012) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2012.6232137 · Cited by 55
“As proposed in [27], Stixels are extracted from a stereo image pair in two steps: the stereo computation, e.g. using semi-global matching stereo (SGM) [12], [15], and the actual Stixel computation.”
BackgroundA survey on human detection surveillance systems for Raspberry Pi
Khalifa, A. F., Badr, E., & Elmahdy, H. N. (2019) · Image and Vision Computing · doi:10.1016/j.imavis.2019.02.010 · Cited by 47
“Other datasets are TUD [39], CVC [40], DaimlerChrysler (DC) [41], and ETH [42].”
MethodPyramid binary pattern features for real-time pedestrian detection from infrared videos
Sun, H., Wang, C., Wang, B., & El‐Sheimy, N. (2010) · Neurocomputing · doi:10.1016/j.neucom.2010.10.009 · Cited by 50
Bilal, M., & Hanif, M. S. (2019) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2019.2906132 · Cited by 45
“The results have been, however, reported for ETH [55] and TUD-Brussels [56] datasets as well to prove the generalization of the proposed methods.”
ResultMulti-pedestrian detection in crowded scenes: A global view
Yan, J., Lei, Z., Yi, D., & Li, S. Z. (2012) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2012.6248045 · Cited by 44
“Further improvements include [19, 6, 5, 10, 20, 18] and so on, where novel features or more powerful classifiers were utilized .”
BackgroundPyramid Center-Symmetric Local Binary/Trinary Patterns for Effective Pedestrian Detection
Zheng, Y., Shen, C., Hartley, R. I., & Huang, X. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-19282-1_23 · Cited by 47
“Springer-Verlag Berlin Heidelberg 2011 1. using discriminative and robust image features, such as Haar wavelets [6], region covariance [10, 12], HOG [8, 9] and PHOG [16]; 2. using a combination of multiple complementary features [14]; 3. including spatial information [16]; 4. the choices of classifiers, such as SVMs […”
MethodA novel data augmentation scheme for pedestrian detection with attribute preserving GAN
Liu, S., Guo, H., Hu, J., Zhao, X. F., Zhao, C., Wang, T., Zhu, Y., Wang, J., & Tang, M. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.02.094 · Cited by 43
Hattori, H., Lee, N., Boddeti, V. N., Beainy, F., Kitani, K., & Kanade, T. (2018) · International Journal of Computer Vision · doi:10.1007/s11263-018-1077-3 · Cited by 44
“This stands in contrast to the large body of work focused primarily on pedestrian detection from a mobile platform [64, 65, 43, 66].”
Result${\rm C}^{4}$: A Real-Time Object Detection Framework
Wu, J., Liu, N., Geyer, C., & Rehg, J. M. (2013) · IEEE Transactions on Image Processing · doi:10.1109/tip.2013.2270111 · Cited by 47
A deep survey on supervised learning based human detection and activity classification methods
Khan, M. A., Mittal, M., Goyal, L. M., & Roy, S. S. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-10811-5 · Cited by 45
“In each round, a weak classifier method is chosen in order to reduced the optimal training error [4, 165, 227, 238, 309, 337, 349].”
MethodPedestrian Detection Using Stationary Wavelet Dilated Residual Super-Resolution
Hsu, W., & Chen, P.-C. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3142061 · Cited by 50
“The database contains 1063 images with shielded pedestrians collected from Caltech [39], ETH [34], TUD-Brussels [40], INRIA [32], Caviar [41], and The Chinese University of Hong Kong.”
MethodJoint Attention in Driver-Pedestrian Interaction: from Theory to PracticePreprint
Rasouli, A., & Tsotsos, J. K. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.02522 · Cited by 44
“5K+ Gr 21K+ BB Daimler-Occ [329] 2010 90K Gr,St 150K+ CL TUD Brussels [330] 2010 3K+ Col 1.”
Fast Pedestrian Detection by Cascaded Random Forest with Dominant Orientation Templates
Tang, D., Liu, Y., & Kim, T. (2012) · British Machine Vision Conference (BMVC) · doi:10.5244/c.26.58 · Cited by 39
“[31] provided an empirical proof that combining different features with HOG can improve the performance drastically.”
BackgroundDisparity Statistics for Pedestrian Detection: Combining Appearance, Motion and Stereo
Walk, S., Konrad Schindler, B., & Schiele, B. (2010) · Lecture notes in computer science · doi:10.1007/978-3-642-15567-3_14 · Cited by 48
“The same features and classifiers were used recently in [4].”
BackgroundMethodInfluentialSelf-citationVisual People Detection - Different Models, Comparison and Discussion
Schiele, B., Andriluka, M., Majer, N., Roth, S., & Wojek, C. (2009) · Cited by 38
Self-citationA review on vision-based pedestrian detection in intelligent transportation systems
Li, B., Yao, Q., & Wang, K. (2012) · IEEE International Conference on Networking, Sensing and Control (ICNSC) · doi:10.1109/icnsc.2012.6204951 · Cited by 40
“[42][46] combine HOG with HOF to detect pedestrians in low resolution images.”
MethodD$^2$-City: A Large-Scale Dashcam Video Dataset of Diverse Traffic Scenarios
Che, Z., Li, G., Li, T., Jiang, B., Shi, X., Zhang, X., Lu, Y., Wu, G., Liu, Y., & Ye, J. (2019) · arXiv · doi:10.48550/arxiv.1904.01975 · Cited by 36
“Traffic datasets in earlier stages were mainly for pedestrian detection and tracking [6, 7, 8, 19].”
BackgroundToprak, T., Belenlioglu, B., Aydin, B., Guzelis, C., & Selver, M. A. (2020) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2020.2983825 · Cited by 43
“These endless challenges motivate creation of datasets with diverse properties (i.e., Caltech [23], INRIA [10], Citypersons [24], KITTI [25], TUD [26], ETH [27], Daimler [7]) to represent real life situations in a better way.”
BackgroundFrom Traditional to Autonomous Vehicles: A Systematic Review of Data Availability
Masello, L., Sheehan, B., Murphy, F., Castignani, G., McDonnell, K., & Ryan, C. (2021) · Transportation Research Record Journal of the Transportation Research Board · doi:10.1177/03611981211057532 · Cited by 37
A Survey of FPGA-Based Vision Systems for Autonomous Cars
Castells‐Rufas, D., Ngo, V., Borrego-Carazo, J., Codina, M., Sanchez, C., Gil, D., & Carrabina, J. (2022) · IEEE Access · doi:10.1109/access.2022.3230282 · Cited by 36
“PD datasets (see table 11) usually use a metric which is based on a miss rate (MR) with respect to pedestrians having a size in pixels larger than a given threshold (typically 50 pixels, corresponding to 60 meters as calculated in [221]).”
Leveraging RGB-D Data: Adaptive fusion and domain adaptation for object detection
Spinello, L., & Arras, K. O. (2012) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2012.6225137 · Cited by 42
“Window-scrolling has also been applied to depth data or other modalities for object detection [31, 26, 8].”
BackgroundMethodInfluentialA Boosted Multi-Task Model for Pedestrian Detection With Occlusion Handling
Zhu, C., & Peng, Y. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2483376 · Cited by 35
Pedestrian Detection for UAVs Using Cascade Classifiers and Saliency Maps
Aguilar, W. G., Luna, M. A., Moya, J. F., Abad, V., Ruiz, H., Parra, H., & Ángulo, C. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-59147-6_48 · Cited by 37
When Pedestrian Detection Meets Nighttime Surveillance: A New Benchmark
Wang, X., Chen, J., Wang, Z., Liu, W., Satoh, S., Liang, C., & Lin, C. (2020) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2020/71 · Cited by 39
“Several datasets have been built for pedestrian detection at daytime, such as INRIA [ Dalal and Triggs, 2005 ] , ETH [ Ess et al. , 2008 ] , TUD-Brussels [ Wojek et al. , 2009 ] , Daim-ler [ Enzweiler and Gavrila, 2009 ] .”
MethodIt’s Not All About Size: On the Role of Data Properties in Pedestrian Detection
Rasouli, A., Kotseruba, I., & Tsotsos, J. K. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-11009-3_12 · Cited by 39
Training Effective Node Classifiers for Cascade Classification
Shen, C., Wang, P., Paisitkriangkrai, S., & van den Hengel, A. (2013) · International Journal of Computer Vision · doi:10.1007/s11263-013-0608-1 · Cited by 35
“Our detector is then compared with existing pedestrian detectors listed in (Dollár et al. 2012), on the INRIA, TUD-Brussels and ETH data sets.”
MethodSIRA: Scale illumination rotation affine invariant mask R-CNN for pedestrian detection
Gawande, U. H., Hajari, K. O., & Golhar, Y. (2022) · Applied Intelligence · doi:10.1007/s10489-021-03073-z · Cited by 33
“The most commonly used hand crafted features are HOG [36], LBP [37], Integral channel features [38], Gray level co-occurrence matrix [47], CNN features [31], HAAR-Wavelet [48], and Oriented gradients [49].”
MethodAn Improved Labelling for the INRIA Person Data Set for Pedestrian Detection
Taiana, M., Nascimento, J. C., & Bernardino, A. J. M. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-38628-2_34 · Cited by 34
“The TUD-MotionPairs/TUD-Brussels data set [14] (TUD) and the Caltech pedestrian data set [2] were introduced in 2009 and contain sequences of images taken in automotive scenarios.”
MethodVulnerable road user detection and safety enhancement: A comprehensive survey
Silva, R. M., Azevedo, G. F., Berto, M. V. V., Rocha, J. R., Fidelis, E. C., Nogueira, M. V., Lisboa, P., & Almeida, T. A. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.128529 · Cited by 34
“In VRU detection, image normalization is a common pre-processing technique that standardizes data fed into deep learning models, enhancing their generalization capabilities [179].”
Vision-Based Human Detection Techniques: A Descriptive Review
Sumit, S. S., Rambli, D. R. A., & Mirjalili, S. (2021) · IEEE Access · doi:10.1109/access.2021.3063028 · Cited by 35
“TUD-Brussels [309] It contains colour images in which 1776 humans, 218 negative images and 1098 positive images are for training while 1498 humans and 508 positive images are for testing.”
BackgroundVision-Based System for Human Detection and Tracking in Indoor Environment
Benezeth, Y., Emile, B., Laurent, H., & Rosenberger, C. (2009) · International Journal of Social Robotics · doi:10.1007/s12369-009-0040-4 · Cited by 31
“[40] propose another approach by using the MPLBoost classifier that simultaneously learns the data partitions and a classifier for each partition.”
MethodEffective Pedestrian Detection Using Center-symmetric Local Binary/Trinary PatternsPreprint
Zheng, Y., Shen, C., Hartley, R. I., & Huang, X. (2010) · arXiv (Cornell University) · doi:10.48550/arxiv.1009.0892 · Cited by 32
“…such as Haar wavelets [6], region covariance [10], [12], HOG [8], [9] and PHOG [20]; 2) using a combination of multiple complementary features [14], [21]; 3) including spatial information [20]; 4) the choices of classifiers, such as support vector machines (SVMs) [8], [20], boosting [22], [23].”
MethodGPU & CPU cooperative accelerated pedestrian and vehicle detection
Machida, T., & Naito, T. (2011) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2011.6130285 · Cited by 35
“Conventional pedestrian detection is more complicated than conventional vehicle detection due to the applied com binations of many sensors and many detection algorithms [8] [10] [12] [14] [16].”
BackgroundPerimeter intrusion detection based on intelligent video analysis
Zhang, Y., Zhang, Z., Xiao, G., Wang, R., & He, X. (2015) · International Conference on Control, Automation and Systems (ICCAS) · doi:10.1109/iccas.2015.7364811 · Cited by 28
Ye, Q., Liang, J., & Jiao, J. (2011) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2011.2167145 · Cited by 28
“1.”
BackgroundGoto, K., Kidono, K., Kimura, Y., & Naito, T. (2011) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2011.5940432 · Cited by 28
“A wide variety of pedestrian detection methods have been proposed [2-6].”
BackgroundSingle Shot Multibox Detector With Kalman Filter for Online Pedestrian Detection in Video
Yang, F., Chen, H., Li, J., Li, F., Wang, L., & Yan, X. (2019) · IEEE Access · doi:10.1109/access.2019.2895376 · Cited by 29
Vision-Based People Detection System for Heavy Machine Applications
Frémont, V., Bui, M. T., Boukerroui, D., & Letort, P. (2016) · Sensors · doi:10.3390/s16010128 · Cited by 27
“Another dataset created at Technische Universitat Darmstadt (TUD) [19] was used to compare different image features and classification methods in real-time applications.”
MethodVideo based human crowd analysis using machine learning: a survey
Chaudhary, D., Kumar, S., & Dhaka, V. S. (2021) · Computer Methods in Biomechanics and Biomedical Engineering Imaging & Visualization · doi:10.1080/21681163.2021.1986859 · Cited by 27
“Wojek et al. (2009) uses motion features for human detection instead of body parts in static images.”
BackgroundPedestrian Detection and Tracking Using HOG and Oriented-LBP Features
Ma, Y., Chen, X., & Chen, G. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-24403-2_15 · Cited by 30
“These motion features are widely used in recent pedestrian detection systems [19, 20].”
People detection in surveillance: classification and evaluation
García‐Martín, Á., & Martínez, J. M. (2015) · IET Computer Vision · doi:10.1049/iet-cvi.2014.0148 · Cited by 33
“Global sequence performance is usually measured in terms of precision – recall (PR) curves [33, 47, 56].”
BackgroundExploiting Target Data to Learn Deep Convolutional Networks for Scene-Adapted Human Detection
Wu, S., Wang, S., Laganière, R., Liu, C., Wong, H., & Xu, Y. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2779271 · Cited by 26
“A number of well-established benchmark datasets are available for building robust human detectors, among which are the INRIA [3], ETH [14], TUD-Brussels [15], and Caltech pedestrian datasets [16].”
BackgroundIntegrating Orientation Cue With EOH-OLBP-Based Multilevel Features for Human Detection
Ma, Y., Deng, L., Chen, X., & Guo, N. (2013) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2013.2268991 · Cited by 26
“In [14], the HOG feature, Haar wavelets, and the oriented histograms of flow are combined in a multiple classifier boosting (MCBoost) algorithm to enhance the efficiency of the HOG-based approach.”
MethodEvidential combination of pedestrian detectors
Xu, P., Davoine, F., & Denœux, T. (2014) · British Machine Vision Conference (BMVC) · doi:10.5244/c.28.2 · Cited by 28
“There exist many pedestrian datasets; INRIA [7], ETH [16], TUD-Brussels [44] and Caltech Pedestrian Detection Benchmark [15] are among the most popular ones.”
Fast moving pedestrian detection based on motion segmentation and new motion features
Zhang, S., Klein, D. A., Bauckhage, C., & Cremers, A. B. (2015) · Multimedia Tools and Applications · doi:10.1007/s11042-015-2571-z · Cited by 25
“Similar to [8], we also use the INRIA person dataset as training dataset for HOG features, while we use the TUD-Brussel dataset [34] for MSD features since this dataset provides motion pair annotations.”
MethodIntensity self similarity features for pedestrian detection in Far-Infrared images
Miron, A., Besbes, B., Rogozan, A., Aïnouz, S., & Bensrhair, A. (2012) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2012.6232227 · Cited by 26
A design framework for hierarchical ensemble of multiple feature extractors and multiple classifiers
Kim, K., Lin, H., Choi, J. Y., & Choi, K. (2015) · Pattern Recognition · doi:10.1016/j.patcog.2015.11.006 · Cited by 27
Improving object localization using macrofeature layout selection
Nam, W., Han, B., & Han, J. H. (2011) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2011.6130467 · Cited by 30
“Our macrofeature layout selection algorithm was evaluated in pedestrian detection with several challenging datasets—INRIA [4], ETH [10], TUD-Brussels [20], and Caltech [9] datasets.”
MethodFast Boosting Based Detection Using Scale Invariant Multimodal Multiresolution Filtered Features
Costea, A. D., Varga, R., & Nedevschi, S. (2017) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2017.112 · Cited by 28
“Several works focused on exploiting these modalities and proposed features based on motion [43] [13] [20]; infrared imagery [27]; depth from LIDAR [38] [28] [19] [34] and depth from stereo [20] [45] [27].”
BackgroundOccluded Pedestrian Detection via Distribution-Based Mutual-Supervised Feature Learning
He, Y., Zhu, C., & Yin, X.-C. (2021) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2021.3094800 · Cited by 24
Variant SemiBoost for Improving Human Detection in Application Scenes
Wu, S., Wong, H., & Wang, S. (2017) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2017.2672686 · Cited by 24
“Based on the well-established evaluation metrics and benchmarks, e.g., INRIA [1], TUD-Brussels [2], Daimler [3], Caltech-USA [4], and KITTI [5], many researchers have taken lots of efforts to significantly improve detection performance in the last two decades.”
BackgroundAn improved fuzzy ARTMAP and Q-learning agent model for pattern classification
Pourpanah, F., Wang, R., Lim, C. P., Wang, X., Seera, M., & Tan, C. J. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.06.002 · Cited by 25
“The results demonstrated the effectiveness of the proposed model in comparison with other methods reported in [39,40] .”
ResultEDGE20: A Cross Spectral Evaluation Dataset for Multiple Surveillance Problems
Le, H. A., Smailis, C. V., Shi, L., & Kakadiaris, I. A. (2020) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv45572.2020.9093573 · Cited by 23
“Year # Images # Pedestrians VIS NIR FIR Thermal Caltech [7] 2009 249K 347K TUD-Brussels [28] 2009 0.5K 1.3K GM-ATCI [25] 2014 137K 200K KAIST [14] 2015 95K 86K CVC [11] 2016 10K 18K CityPersons [31] 2017 5K 35K SCUT [29] 2018 Section 4 presents statistics of the EDGE20 dataset and the evaluation…”
YOLOpeds: efficient real‐time single‐shot pedestrian detection for smart camera applications
Kyrkou, C. (2020) · IET Computer Vision · doi:10.1049/iet-cvi.2019.0897 · Cited by 24
Pedestrian Counting Using Yolo V3
Menon, A., Omman, B., & Asha, S. B. (2021) · International Conference on Innovative Trends in Information Technology (ICITIIT) · doi:10.1109/icitiit51526.2021.9399607 · Cited by 24
“Evaluated the performance of the DW SPLS approach on three challenging human datasets, including INRIA [35], Caltech [36], and TUD-Brussels [37] pedestrian data sets.”
MethodIntegrated Pedestrian and Direction Classification Using a Random Decision Forest
Tao, J., & Klette, R. (2013) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2013.38 · Cited by 25
“The derivation of global or holistic descriptors for a rectangular bounding box with a trained classifier is a standard approach in the field [7, 8, 9, 10, 17, 27, 28, 24].”
BackgroundMulti-spectral pedestrian detection
Yuan, Y., Lu, X., & Chen, X. (2014) · Signal Processing · doi:10.1016/j.sigpro.2014.08.003 · Cited by 22
“For pedestrian datasets of visible light, there are several popular datasets including the INRIA [7], Caltech [20], ETH [21] and TUD-Brussels [22].”
MethodDeep Learning Approach to Key Frame Detection in Human Action Videos
Gawande, U. H., Hajari, K. O., & Golhar, Y. (2020) · IntechOpen eBooks · doi:10.5772/intechopen.91188 · Cited by 22
Xie, H., Chen, Y., & Shin, H. (2018) · Applied Intelligence · doi:10.1007/s10489-018-1326-8 · Cited by 24
“DNN + SS [2], Caltech training set, ETH [25] and TudBrussels dataset [26] are used.”
MethodMultiple Viewpoint Recognition and Localization
Helmer, S., Meger, D., Muja, M., Little, J. J., & Lowe, D. G. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-19315-6_36 · Cited by 24
“…for outdoor navigation such as intelligent driving systems (see several recent vision based methods among far too many others too mention [10, 11]), however in most cases these systems have attempted to recognize just a few of the most relevant object categories for the task of safe…”
Efficient object detection and segmentation with a cascaded Hough Forest ISM
Rematas, K., & Leibe, B. (2011) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2011.6130356 · Cited by 23
“Performance of our approach and comparison of the proposed framework against [26, 7] on the ETH Person dataset.”
BackgroundMethodInfluentialYOLOv2PD: An Efficient Pedestrian Detection Algorithm Using Improved YOLOv2 Model
Chintakindi, B. M., Hashmi, M. F., Muhammad, G., & AlQahtani, S. A. (2021) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2021.018781 · Cited by 22
“The ConvNet [27] method uses convents for detecting pedestrians.”
MethodImproving pedestrian detection with selective gradient self-similarity feature
Wu, S., Laganière, R., & Payeur, P. (2015) · Pattern Recognition · doi:10.1016/j.patcog.2015.01.005 · Cited by 20
“Dataset The test dataset includes the INRIA [3], ETH [37], TUD-Brussels [5], and Caltech [38] pedestrian datasets.”
BackgroundMethodInfluentialDeep Learning-Based Multi-scale Multi-object Detection and Classification for Autonomous Driving
Fink, M., Liu, Y., Engstle, A., & Schneider, S.-A. (2019) · Proceedings · doi:10.1007/978-3-658-23751-6_20 · Cited by 20
The WILDTRACK Multi-Camera Person DatasetPreprint
Chavdarova, T., Baqué, P., Bouquet, S., Maksai, A., Jose, C., Lettry, L., Fua, P., Gool, L. V., & Fleuret, F. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1707.09299 · Cited by 27
Saeedizadeh, N., Jalali, S. M. J., Khan, B., & Mohamed, S. M. K. (2025) · Expert Systems · doi:10.1111/exsy.70020 · Cited by 21
“A driving car in the inner city of Brussels recorded the new TUD –Brussels dataset (Wojek et al. 2009; Ouyang et al. 2014).”
Cascade Classifiers and Saliency Maps Based People Detection
Aguilar, W. G., Luna, M. A., Moya, J. F., Abad, V., Ruiz, H., Parra, H., & Lopez, W. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-60928-7_42 · Cited by 22
IRUVD: a new still-image based dataset for automatic vehicle detection
Ali, A., Sarkar, R., & Das, D. K. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-15365-2 · Cited by 19
“For the sole purpose of detecting pedestrians, datasets like INRIA [7], Daimler Pedestrian [27], TudBrussels [43], and Citypersons dataset [45] were created.”
An Evaluation of the Pedestrian Classification in a Multi-Domain Multi-Modality Setup
Miron, A., Rogozan, A., Aïnouz, S., Bensrhair, A., & Broggi, A. (2015) · Sensors · doi:10.3390/s150613851 · Cited by 20
“In [19], motion estimation is incorporated, using HOG, Haar and HOF.”
MethodOn collaborative people detection and tracking in complex scenarios
García‐Martín, Á., & Martínez, J. M. (2012) · Image and Vision Computing · doi:10.1016/j.imavis.2012.03.005 · Cited by 19
“Over the last few years, there have been multiple approaches in more realistic environments with multiple people and occlusions [3] and even onboard scenarios [4].”
BackgroundAuto++: Detecting Cars Using Embedded Microphones in Real-Time
Li, S., Fan, X., Zhang, Y., Trappe, W. K., Lindqvist, J., & Howard, R. E. (2017) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3130938 · Cited by 34
“A large number of such studies have been conducted for users when they are driving [3, 30, 35], while much less effort has been devoted to users when they are walking.”
BackgroundMoctezuma, D., Conde, C., de Diego, I. M., & Cabello, E. (2011) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2011.6130466 · Cited by 20
“Other evaluation of different features and classifiers is presented in [27].”
BackgroundDifferential Features for Pedestrian Detection: A Taylor Series Perspective
Shen, J., Zuo, X., Yang, W., Prokhorov, D. V., Mei, X., & Ling, H. (2018) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2018.2869087 · Cited by 18
“1http://www.vision.caltech.edu/Image_Datasets/CaltechPedestrians/ 2) TUD-Brussel Dataset: The experimental results on the TUD-Brussel dataset are illustrated in Fig.”
MethodInfluentialOtto, C., Gerber, W., León, F. P., & Wirnitzer, J. (2012) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2012.6232228 · Cited by 20
“A subsequent non-maximum suppression algorithm [16] enhances the detection performance and provides another value of certainty between −1 and 1.”
BackgroundDriving Datasets Literature ReviewPreprint
Laflamme, C.-É. N., Pomerleau, F., & Giguère, P. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.11968 · Cited by 21
Vision-based pedestrian behavior analysis at intersections
Shirazi, M. S., & Morris, B. (2016) · Journal of Electronic Imaging · doi:10.1117/1.jei.25.5.051203 · Cited by 16
Discriminative latent semantic feature learning for pedestrian detection
Zhu, C., & Peng, Y. (2017) · Neurocomputing · doi:10.1016/j.neucom.2017.01.043 · Cited by 16
Opponent Colors for Human Detection
Anwer, R. M., Vázquez, D., & López, A. M. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-21257-4_45 · Cited by 18
“as well as in many following works [4, 11–15], compute HOG on top of RGB space.”
MethodAggregated channels network for real-time pedestrian detection
Ghorban, F., Marín, J., Su, Y., Colombo, A., & Kummert, A. (2018) · International Conference on Machine Vision · doi:10.1117/12.2309864 · Cited by 16
“Decision forests [2, 11, 25, 37, 44], deformable parts models (DPM) [12, 13, 27, 29, 39], and deep networks [1, 4, 19, 22, 38, 41].”
BackgroundDepth sensor based human detection for indoor surveillance
Hu, T., Zhang, H., Zhu, X., Clunis, J., & Yang, G. (2018) · Future Generation Computer Systems · doi:10.1016/j.future.2018.05.083 · Cited by 16
Quality-adaptive deep learning for pedestrian detection
Tahboub, K., Güera, D., Reibman, A. R., & Delp, E. J. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8297071 · Cited by 15
“Some of the most popular benchmarks include: INRIA person dataset [19], ETH dataet [24], TUD-Brussels pedestrian dataset [25], Daimler detection benchmark (Daimler-DB) [26] and Caltech pedestrian dataset [27].”
BackgroundReinforced Bit Allocation under Task-Driven Semantic Distortion Metrics
Shi, J., & Chen, Z. (2020) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas45731.2020.9180454 · Cited by 15
Pedestrian Detection with Semantic Regions of Interest
He, M., Luo, H., Chang, Z., & Hui, B. (2017) · Sensors · doi:10.3390/s17112699 · Cited by 17
“The TUD-Brussels Pedestrian Dataset is a dataset with image pairs recorded in a crowded urban setting with an onboard camera [35].”
MethodA heuristic SVM based pedestrian detection approach employing shape and texture descriptors
Kumar, K., & Mishra, R. K. (2020) · Multimedia Tools and Applications · doi:10.1007/s11042-020-08864-z · Cited by 16
“A number of pedestrian datasets, such as INRIA [4], Daimler [7], TUD-Brussels-Motion Pairs [31], ETH [8] are available for training and testing the performance of pedestrian detectors in various environments.”
BackgroundGuo, J., Kurup, U., & Shah, M. (2018) · arXiv · doi:10.48550/arxiv.1811.11277 · Cited by 14
“2011; Northern Italy X X X X X 28 VI TUD-Brussels Pedestrian [111] Max Planck Inst for Informatics 2009; Belgium X X X 1.”
BackgroundCRNN Based Jersey-Bib Number/Text Recognition in Sports and Marathon Images
Nag, S., Ramachandra, R., Shivakumara, P., Pal, U., Lü, T., & Kankanhalli, M. (2019) · IAPR International Conference on Document Analysis and Recognition (ICDAR) · doi:10.1109/icdar.2019.00186 · Cited by 19
“Training Details: To train the SSD, we have used all the images which contain at least one annotated pedestrian from the training set of Caltech pedestrian dataset [21], TUDBrussels dataset [22] and INRIA person dataset [23].”
MethodA review on vision-based pedestrian detection
Zheng, G., & Chen, Y. (2012) · IEEE Global High Tech Congress on Electronics (GHTCE) · doi:10.1109/ghtce.2012.6490122 · Cited by 20
“PEDESTRIAN DETECTION DATABASE Commonly used pedestrian detection databases are CAVIAR, INRIA, MIT, CALTECH [65] and TUD-Brussels [66], among which CALTECH and TUD-Brussels are the most challenging ones, as listed in Table.”
MethodInfluentialPedestrian detection based on the privileged information
Meng, F., Qi, Z., Tian, Y., & Niu, L. (2016) · Neural Computing and Applications · doi:10.1007/s00521-016-2639-3 · Cited by 14
C-RPNs: Promoting object detection in real world via a cascade structure of Region Proposal Networks
Yang, D., Zou, Y., Zhang, J., & Li, G. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.08.016 · Cited by 14
“Following the protocol in [40], one frame from every five frames of Caltech Benchmark and all frames of the ETH [41] and TUD-Brussels [42] are extracted as training data, which includes 27,021 images in total.”
MethodPeople detection based on appearance and motion models
García‐Martín, Á., Hauptmann, A. G., & Martínez, J. M. (2011) · IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) · doi:10.1109/avss.2011.6027333 · Cited by 13
“Over the last few years, there have been multiple approaches in more realistic environments with multiple people and occlusions [2, 12], and even onboard scenarios [17].”
BackgroundCognitive pedestrian detector: Adapting detector to specific scene by transferring attributes
Zhang, X., He, F., Tian, L., & Wang, S. (2014) · Neurocomputing · doi:10.1016/j.neucom.2014.07.054 · Cited by 13
Real time person detection and tracking by mobile robots using RGB-D images
Vo, D. M., Jiang, L., & Zell, A. (2014) · IEEE International Conference on Robotics and Biomimetics (ROBIO) · doi:10.1109/robio.2014.7090411 · Cited by 14
“State-of-the-art algorithms of human detection [6], [7], make a great contribution to tracking-by-detection approaches [8], [9], thus significantly improving the tracking system of mobile robots.”
BackgroundBartoli, F., Lisanti, G., Seidenari, L., Karaman, S., & Bimbo, A. D. (2015) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2015.7301279 · Cited by 13
“However, this dataset largely differs from the classic pedestrian detection datasets [11, 16] since it was mainly designed for social behavior analysis on single-shot images.”
MethodDriving Space for Autonomous Vehicles
Yang, D., Jiao, X., Jiang, K., & Cao, Z. L. (2019) · Automotive Innovation · doi:10.1007/s42154-019-00081-1 · Cited by 16
SPID: Surveillance Pedestrian Image Dataset and Performance Evaluation for Pedestrian Detection
Wang, D., Zhang, C., Cheng, H., Shang, Y., & Mei, L. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-54526-4_34 · Cited by 15
“Multiple public pedestrian datasets have been collected over the past decades, including commonly used INRIA, ETH [2], TUD-Brussels [5], Daimler [6], Caltech and KITTI datasets [4].”
Method
Metadata
- Authors
- Wojek, C., Walk, S., & Schiele, B.
- Affiliations
- Technische Universität Darmstadt (DE)
- ORCID
- Schiele, Bernt
- Venue
- 2009 IEEE Conference on Computer Vision and Pattern Recognition
- Publisher
- IEEE
- Published
- 2009-06
- Pages
- 794-801
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W2127420331
- Cited by
- 350
- References
- 37
- Citations by year
- 2026: 2 · 2025: 5 · 2024: 1 · 2023: 8 · 2022: 13 · 2021: 21 · 2020: 24 · 2019: 25 · 2018: 27 · 2017: 30 · 2016: 24 · 2015: 30 · 2014: 29 · 2013: 37 · 2012: 19
- FWCI
- 20.14
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- Max Planck Digital Library (submittedVersion)
- MPG.PuRe (Max Planck Society) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Video Surveillance and Tracking Methods
- Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- pedestrian detection, onboard camera, motion information, multi-cue fusion, feature combination, sliding window, classifier comparison, urban driving
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.