Detector ensemble based on false positive mining for pedestrian detection (original) (raw)

Pedestrian Detection in Automotive Safety: Understanding State-of-the-Art

Pedestrian detection is one of the important computer vision problems in automotive safety and driver assistance domain. It is a major component of the advanced driver assistance system (ADAS) which help the driver to drive safely. Recent literature shows a number of research activities addressing object detection/tracking in general and pedestrian detection in particular. The solutions proposed by different researchers vary in detection methods, detection scenario, feature descriptors, classification schemes, detection performance, as well as computational complexity. However, the average detection accuracy is not much promising even after many years of research. The fail-safe and real-time human detection from real life road scenes, even in standard resolution, is far from reality. Safety critical systems in the automotive industry have to follow well established stringent safety standards like ISO26262. Since the pedestrian detection system deals with human safety, it also has to follow these standards before integrating to the vehicle electronics. This paper is a study of different techniques used in pedestrian detection specific to the automotive application, along with a description of generic pedestrian detection solution architecture. INDEX TERMS Road safety, advanced driver assistance systems (ADAS), smart image region of interest (ROI) selection, image feature descriptors, pedestrian detection.

Is pedestrian detection really a hard task

2007

Abstract In this paper we present a simple approach for person detection in surveillance for static cameras. The basic idea is to train a separate classifier for each image location which has only to discriminate the object from the background at a specific location. This is a considerably simpler problem than the detection of persons on arbitrary backgrounds. Therefore, we use adaptive classifiers which are trained online. Due to the reduced complexity we can use a simple update strategy that requires only a few positive samples ...

Pedestrian Detection: The Elephant In The Room

2020

Pedestrian detection is used in many vision based applications ranging from video surveillance to autonomous driving. Despite achieving high performance, it is still largely unknown how well existing detectors generalize to unseen data. To this end, we conduct a comprehensive study in this paper, using a general principle of direct cross-dataset evaluation. Through this study, we find that existing state-of-the-art pedestrian detectors generalize poorly from one dataset to another. We demonstrate that there are two reasons for this trend. Firstly, they over-fit on popular datasets in a traditional single-dataset training and test pipeline. Secondly, the training source is generally not dense in pedestrians and diverse in scenarios. Accordingly, through experiments we find that a general purpose object detector works better in direct cross-dataset evaluation compared with state-of-the-art pedestrian detectors and we illustrate that diverse and dense datasets, collected by crawling th...

Bounding Box Accuracy in Pedestrian Detection for Intelligent Transportation Systems

2006

This paper describes a stereo-vision-based pedestrian detection system for intelligent transportation systems. The basic components of pedestrians are first located in the image and then combined with a SVM-based classifier. Generic obstacles are located using a subtractive clustering attention mechanism based on stereo vision. A by-components learning approach is proposed and different feature extraction methods are tested in order to better deal with pedestrian variability and justify what features are better to be learnt for pedestrian detection. Candidate selection mechanisms usually yield pedestrians with inaccurate bounding boxes. Then a decrease in detection rate takes place if the SVM classifier is trained only with well-fitted pedestrians. Using several off-line databases containing thousands of pedestrians samples the effect of bounding box accuracy is studied. A multi-candidate generation mechanism is also developed in order to enhance the single frame performance, decreasing the number of false positives due to inaccurate bounding boxes

On exploration of classifier ensemble synergism in pedestrian detection

… Transportation Systems, IEEE …, 2010

A single feature extractor-classifier is not usually able to deal with the diversity of multiple image scenarios. Therefore, integration of features and classifiers can bring benefits to cope with this problem, particularly when the parts are carefully chosen and synergistically combined. In this paper, we address the problem of pedestrian detection by a novel ensemble method. Initially, histograms of oriented gradients (HOGs) and local receptive fields (LRFs), which are provided by a convolutional neural network, have been both classified by multilayer perceptrons (MLPs) and support vector machines (SVMs). A diversity measure is used to refine the initial set of feature extractors and classifiers. A final classifier ensemble was then structured by an HOG and an LRF as features, classified by two SVMs and one MLP. We have analyzed the following two classes of fusion methods of combining the outputs of the component classifiers: 1) majority vote and 2) fuzzy integral. The first part of the performance evaluation consisted of running the final proposed ensemble over the DaimlerChrysler cropwise data set, which was also artificially modified to simulate sunny and shadowy illumination conditions, which is typical of outdoor scenarios. Then, a window-wise study has been performed over a collected video sequence. Experiments have highlighted a state-of-the-art classification system, performing consistently better than the component classifiers and other methods.

Survey of pedestrian detection for advanced driver assistance systems

Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2010

Advanced driver assistance systems (ADASs), and particularly pedestrian protection systems (PPSs), have become an active research area aimed at improving traffic safety. The major challenge of PPSs is the development of reliable on-board pedestrian detection systems. Due to the varying appearance of pedestrians (eg, different clothes, changing size, aspect ratio, and dynamic shape) and the unstructured environment, it is very difficult to cope with the demanded robustness of this kind of system.

Pedestrian Detection with Minimal False Positives per Color-Thermal Image

This research is based on aggregate channel features utilized for pedestrian detection, and the main focus is to investigate a simple way to reduce the number of false positives per image. The importance of this will be to increase the accuracy of the detector by removing the excessive number of false positives while maintaining the missing rate as low as possible. To omit such unwanted false positives, we utilized an image categorization method for day and night images in order to minimize the misclassification rate. Furthermore, the best extension of the aggregate channel features method was analyzed and is recommended as a base detector. As a result, a night-time pre-trained pedestrian detector is only applied to night images, and a daytime detector is applied to daytime images. Thus, a large number of false positives are avoided while the missing rate is greatly reduced.

Vision Based Pedestrian Detection for Advanced Driver assistance

Procedia Computer Science, 2015

In pedestrian detection intricate feature descriptors are used to improve the detection rate at the cost of computational complexity. In this paper, we propose a detector based on simple, robust edgelet features to enhance the detection rate and classifier based on k-means clustering approach to reduce computational complexity. The proposed framework consists of extraction of candidate features of pedestrian detection using edgelet features and use of the cascade structure of k-means clustering for classification enabling high detection accuracy at low false positives. Experimental results show that the proposed method requires less processing time per frame, making it suitable for real-time systems.

A Robust Pedestrian Detection Approach for Autonomous Vehicles

2022 8th Iranian Conference on Signal Processing and Intelligent Systems (ICSPIS)

Nowadays, utilizing Advanced Driver-Assistance Systems (ADAS) has absorbed a huge interest as a potential solution for reducing road traffic issues. Despite recent technological advances in such systems, there are still many inquiries that need to be overcome. For instance, ADAS requires accurate and real-time detection of pedestrians in various driving scenarios. To solve the mentioned problem, this paper aims to fine-tune the YOLOv5s framework for handling pedestrian detection challenges on the real-world instances of Caltech pedestrian dataset. We also introduce a developed toolbox for preparing training and test data and annotations of Caltech pedestrian dataset into the format recognizable by YOLOv5. Experimental results of utilizing our approach show that the mean Average Precision (mAP) of our fine-tuned model for pedestrian detection task is more than 91 percent when performing at the highest rate of 70 FPS. Moreover, the experiments on the Caltech pedestrian dataset samples have verified that our proposed approach is an effective and accurate method for pedestrian detection and can outperform other existing methodologies.