Ten Years of Pedestrian Detection, What Have We Learned?

Benenson, Rodrigo; Omran, Mahamed G. H.; Hosang, Jan; Schiele, Bernt

doi:10.1007/978-3-319-16181-5_47

Cited by 424 publications

(393 citation statements)

References 57 publications

Supporting

Mentioning

383

Contrasting

Unclassified

Order By: Relevance

“…1 Examples of where occlusion arises in a single-camera system, showing (a) almost complete occlusion and (b) partial occlusion. In such scenarios, the person in the foreground has their skeleton detected, while the person being occluded is ignored and color channels, along with discrete cosine transforms [18]. The best results are achieved with one of the three machine learning algorithms: deformable part models, convolutional neural networks, and decision forests [18].…”

Section: Related Workmentioning

confidence: 90%

Out of sight: a toolkit for tracking occluded human joint positions

Quigley

Harris-Birtill

2016

Pers Ubiquit Comput

View full text Add to dashboard Cite

Real-time identification and tracking of the joint positions of people can be achieved with off-the-shelf sensing technologies such as the Microsoft Kinect, or other camera-based systems with computer vision. However, tracking is constrained by the system's field of view of people. When a person is occluded from the camera view, their position can no longer be followed. Out of Sight addresses the occlusion problem in depth-sensing tracking systems. Our new tracking infrastructure provides human skeleton joint positions during occlusion, by combining the field of view of multiple Kinects using geometric calibration and affine transformation. We verified the technique's accuracy through a system evaluation consisting of 20 participants in stationary position and in motion, with two Kinects positioned parallel, 45, and 90 apart. Results show that our skeleton matching is accurate to within 16.1 cm (s.d. = 5.8 cm), which is within a person's personal space. In a realistic scenario study, groups of two people quickly occlude each other, and occlusion is resolved for 85% of the participants. A RESTful API was developed to allow distributed access of occlusion-free skeleton joint positions. As a further contribution, we provide the system as open source.

show abstract

Section: Related Workmentioning

confidence: 90%

Out of sight: a toolkit for tracking occluded human joint positions

Quigley

Harris-Birtill

2016

Pers Ubiquit Comput

View full text Add to dashboard Cite

show abstract

“…Decision forests are obtained by combining the predictions of multiple independent Decision Tree Models to obtain a single prediction. Output of ANN and decision forest strategy has been found to be at par with each other [22]. However greater complexity incurred in these approaches, advocates the usage of SVM, which gives comparable performance with a much smaller implementation complexity.…”

Section: Literature Surveymentioning

confidence: 99%

“…The histogram comprises of nine bins obtained by equally dividing the interval 0-180 o into bins of interval 20 o .Consider the 8x8 cell region as shown in Figure 4. Assuming that the pixel (1,1) corresponds to a gradient orientation value of 30 o , it can be mapped to the bin corresponding to the interval [21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40] o .In other words, the gradient magnitude corresponding to the orientation value of 30 o , will be added to the histogram bin of 21-40 o for that pixel. Similarly, the next pixel (1,2) with a gradient orientation value of 55 o , is mapped with its gradient magnitude into the bin corresponding to [41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56][57][58][59][60] o in the histogram of the cell under consideration.…”

Section: Spatial / Orientation Binningmentioning

confidence: 99%

Implementation of Robust HOG-SVM based Pedestrian Classification

Yadav¹,

Senthamilarasu²,

Kutty³

et al. 2015

IJCA

View full text Add to dashboard Cite

Achieving pedestrian protection by means of computer vision is not a new topic in the field of computer vision research; however it is still being pursued with renewed interest because of the huge scope for performance improvement in the existing systems. Generally, the task of pedestrian detection (PD) involves stages such as pre-processing, ROI selection, feature extraction, classification, verification/refinement and tracking. Of all the steps involved in the PD framework, the paper presents the work done towards implementing the feature extraction and classification stages in particular. It is of paramount importance that the extracted features from the image should be robust and distinct enough to help the classifier distinguish between a pedestrian and a non-pedestrian, while a good classification algorithm would go a long way in precisely identifying a pedestrian as well as in simplifying the verification stage of the PD framework. The presented work focuses on the implementation of the Histogram of Oriented Gradients (HOG) features with modified parameters that can represent accurate intrinsic information of the image. Classification is achieved using Support Vector Machine (SVM). However instead of employing a readily available SVM library, the linear SVM implemented uses the Sequential Minimal Optimization (SMO) algorithm. The results observed by this HOG-SVM combination show promise to be the best feature extraction cum classification module for a full-fledged PD system. General TermsComputer Vision, Image processing.

show abstract

“…

…”

mentioning

confidence: 99%

Real-time Human Detection based on Personness Estimation

Kim

Sohn

2015

Procedings of the British Machine Vision Conference 2015

View full text Add to dashboard Cite

Even though deformable part model (DPM) [3] is one of the most popular human detection methods [1], it is difficult to see a practical application of DPM human detection on mobile devices due to its computational overhead. On the other hand, many objectness (a.k.a. detection proposals) measure methods have been recently proposed [4]. An objectness measure generates object windows that are likely to contain generic objects, and allows avoiding an exhaustive search. Its intention to improve detection speed seems to be perfectly matched with DPM human detection in mobile devices. However, most objectness measure methods reviewed and evaluated in [4] are not well suited for real-time detection because all objectness measure methods (except for the binarized normed gradients (BING) method [2]) spend more than 250 milliseconds.For this reason, we propose a more efficient and accurate method for estimating human windows using the normed gradients with color feature in order to enhance detection rate within a short period of time. In this paper, we are interested in humans, not generic objects. So we name objectness estimation for humans simply as personness estimation.To rapidly determine the priority order of each feature vector of an image with a supervised approach, we make use of the skin color model and the fast NG feature (or simply Bing) proposed by Cheng et al. [2]. An NG feature is a 64 dimensional vector describing the magnitude (a.k.a. norm) of the gradients of an 8 × 8 downsampled image. To examine every image window for the objectness measure, a magnitude map of gradients is resized to 36 predefined quantized shapes (or quantized sizes in [2]). The results are called NG maps. By calculating the correlation values between the NG maps and w ∈ R 8×8 , we can obtain a filter score for each window. A quantized shape is ignored in the predicting stage if the quantized shape has less than or equal to 50 positive samples in the training stage. The final objectness score is calculated as follows:where v i ,t i ∈ R and s (i,x,y) are learned coefficient, a bias term and the filter score at position (x, y) of each quantized shape i. s (i,x,y) is obtained using the first stage linear SVM, and v i and t i are obtained using the second stage linear SVM. Even though SVM training on various object categories makes the original linear model w ∈ R 8×8 work for generic objects, it degrades the performance of the SVM classifier for single-category object detection. Therefore, we generate the new linear model w p ∈ R 8×8 shown in Fig. 1(a) by training the linear SVM only on humans that our DPM detector can detect. This restricted training technique also reduces the number of quantized shapes to consider. Fig. 1(a) illustrates that w p ∈ R 8×8 surprisingly places more confidence in the shoulder and head regions. After training the linear SVM on VOC 2007 dataset, we choose four points (3, 1), (3, 2), (4, 1), (4, 2) to extract skin color information. Head and neck are usually located in these four points as shown in the Fig.

show abstract

Ten Years of Pedestrian Detection, What Have We Learned?

Cited by 424 publications

References 57 publications

Out of sight: a toolkit for tracking occluded human joint positions

Out of sight: a toolkit for tracking occluded human joint positions

Implementation of Robust HOG-SVM based Pedestrian Classification

Real-time Human Detection based on Personness Estimation

Contact Info

Product

Resources

About