Most multiple object tracking methods rely on object detection methods in order to initialize new tracks and to update existing tracks. Although strongly interconnected, tracking and detection are usually addressed as separate building blocks. However both parts can benefit from each other, e.g. the affinity model from the tracking method can reuse appearance features already calculated by the detector, and the detector can use object information from past in order to avoid missed detection. Towards this end, we propose a multiple object tracking method that jointly performs detection and tracking in a single neural network architecture. By training both parts together, we can use optimized parameters instead of heuristic decisions over the track lifetime. We adapt the Single Shot MultiBox Detector (SSD) [14] to serve single frame detection to a recurrent neural network (RNN), which combines detections into tracks. We show initial prove of concept on the DETRAC[26] benchmark with competitive results, illustrating the feasibility of learnable track management. We conclude with a discussion of open problems on the MOT16[15] benchmark.
Annotating data for training a person detector is a tedious procedure. Therefore it is worthwhile to use freely available datasets. When detecting in the infrared spectrum it is not obvious that person images from the visible spectrum can be used to train a detector operable in IR. We show that it is possible to train a transmodel detector, which can be used to detect in IR as well as in the visible spectrum. Therefor we use integral channel features in combination with boosting based feature selection, in order to analyze which features are effective for generating the effect of transmodality
<p><strong>Abstract.</strong> In this paper we present a versatile multi-sensor vehicle which is used in several research projects. The vehicle is equipped with various sensors in order to cover the needs of different research projects in the area of object detection and tracking, mobile mapping and change detection. We show an example for the capabilities of this vehicle by presenting camera- and LiDAR-based pedestrian detection methods. Besides this specific use case, we provide a more general in-depth description of the vehicle’s hard- and software design and its data-processing capabilities. The vehicle can be used as a sensor carrier for mobile mapping, but it also offers hardware and software components to allow for an adaptable onboard processing. This enables the development and testing of methods related to real-time applications or high-level driver assistance functions. The vehicle’s hardware and software layout result from several years of experience, and our lessons learned can help other researchers set up their own experimental platform.</p>
Methods for automated person detection and person tracking are essential core components in modern security and surveillance systems. Most state-of-the-art person detectors follow a statistical approach, where prototypical appearances of persons are learned from training samples with known class labels. Selecting appropriate learning samples has a significant impact on the quality of the generated person detectors. For example, training a classifier on a rigid body model using training samples with strong pose variations is in general not effective, irrespective of the classifiers capabilities. Generation of high-quality training data is, apart from performance issues, a very time consuming process, comprising a significant amount of manual work. Furthermore, due to inevitable limitations of freely available training data, corresponding classifiers are not always transferable to a given sensor and are only applicable in a well-defined narrow variety of scenes and camera setups. Semi-supervised learning methods are a commonly used alternative to supervised training, in general requiring only few labeled samples. However, as a drawback semi-supervised methods always include a generative component, which is known to be difficult to learn. Therefore, automated processes for generating training data sets for supervised methods are needed. Such approaches could either help to better adjust classifiers to respective hardware, or serve as a complement to existing data sets. Towards this end, this paper provides some insights into the quality requirements of automatically generated training data for supervised learning methods. Assuming a static camera, labels are generated based on motion detection by background subtraction with respect to weak constraints on the enclosing bounding box of the motion blobs. Since this labeling method consists of standard components, we illustrate the effectiveness by adapting a person detector to cameras of a sensor network. While varying the training data and keeping the detection framework identical, we derive statements about the sample quality
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.