Pedestrian Detection: An Evaluation of the State of the Art

By Piotr Doll'ar, Christian Wojek, Bernt Schiele, Pietro Perona. 2012

3,350 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Microsoft COCO: Common Objects in Context (2014, background): “Many object detection algorithms benefit from additional annotations, such as the amount an instance is occluded [4] or the location of keypoints on the object [10].”
Cascade R-CNN: High Quality Object Detection and Instance Segmentation (2019, neutral): “shold u is used to determine whether a detection is a success (IoU(b,g) ≥ u) or failure (IoU(b,g) < u). Many object detection datasets, including PASCAL VOC [13], ImageNet [48], Caltech Pedestrian [11], etc., use u = 0.5. This is partly because these datasets were established a while ago, when ob”
Deep Perm-Set Net: Learn to predict sets with unknown permutation and cardinality using deep neural networks (2018, methodology): “To quantify the detection performance, we adopt the commonly used evaluation curves and metrics Dollár et al. (2012) such as ROC, precision-recall curves, average precision (AP) and the log-average miss rate (MR) over false positive per image.”
Panoptic Segmentation (2018, background): “Later, pedestrian detection datasets [8] helped drive progress in the field.”
DeepSetNet: Predicting Sets with Deep Neural Networks (2016, methodology): “We perform experiments on two widely used datasets, Caltech Pedestrians [5] and MOT16, the 2016 edition of the MOTChallenge benchmark [22].”
The Cityscapes Dataset for Semantic Urban Scene Understanding (2016, neutral): “We additionally compare to the Caltech Pedestrian Dataset [12], which only contains annotations for humans, but none for vehicles.”
LVIS: A Dataset for Large Vocabulary Instance Segmentation (2019, neutral): “LVIS is inspired by these and other related datasets, including those focused on street scenes (Cityscapes [3] and Mapillary [22]) and pedestrians (Caltech Pedestrians [4]).”
Real-World Image Datasets for Federated Learning (2019, background): “Object detection is at the core of many real-world artiﬁcial intelligence (AI) applications, such as face detection [Sung and Poggio, 1998], pedestrian detection [Dollar et al. , 2012], safety controls, and video analysis.”
A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection (2016, methodology): “Augmentation In [4,6], it is argued that multi-scale training is not needed, since deep neural networks are adept at learning scale invariance. This, however, is not true for datasets such as Caltech [13] and KITTI [10], where object scales can span multiple octaves. In KITTI, many objects are quite”
Visual Causal Feature Learning (2014, background): “Despite very encouraging results in image classification (Krizhevsky et al., 2012), object detection (Dollar et al., 2012) and fine-grained classification (Branson et al., 2014; Zhang et al., 2014), some researchers have found that visual neural networks can be easily fooled using adversarial…”
nuScenes: A Multimodal Dataset for Autonomous Driving (2019, background): “Other datasets focus exclusively on pedestrian annotations on images [20, 25, 79, 24, 88, 23, 58].”
Deep Feature Pyramid Reconfiguration for Object Detection (2018, methodology): “The sliding window methods over multi-scale feature pyramids are usually applied in object detection [10,14].”
High-Level Semantic Feature Detection: A New Perspective for Pedestrian Detection (2019, methodology): “For Caltech [9], a mini-batch contains 16 images with one GPU (GTX 1080Ti), the learning rate is set as 10 and training is stopped after 15K iterations.”
CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes (2018, background): “Most of the early researches focus on detection-based approaches using a moving-window-like detector to detect people and count their number [26].”
Fighting against COVID-19: A novel deep learning model based on YOLO-v2 with ResNet-50 for medical face mask detection (2020, neutral): “At all recall levels, the ideal precision is one. log-average miss rates are a mean of nine False Positives Per Image (FPPI) miss rate in the range of 10 (cid:0) 3 to 10 0 to give stable performance (Dollar, Wojek, Schiele, & Perona, 2012; Dollar, Appel, Belongie, & Perona, 2014).”
YOLO-based Object Detection Models: A Review and its Applications (2024, neutral): “Other datasets that are used in previous research work include Remote sensing target detection [42] [43] [44], text detection [45] [46] [47], and Pedestrian detection [48] [49] [50].”
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art (2017, background): “In their survey on pedestrian detection systems from monocular images, Dollar et al. (2011) present an extensive evaluation focusing on the evaluation of sliding window approaches.”