Fraunhofer Institute for Communication, Information Processing and Ergonomics

Germany

Fraunhofer Institute for Communication, Information Processing and Ergonomics

Germany
SEARCH FILTERS
Time filter
Source Type

Baggenstoss P.M.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
IEEE Transactions on Signal Processing | Year: 2017

Maximum entropy PDF projection (MEPP) is a way to construct generative models from feature transformations. Corresponding to each dimension-reducing feature mapping, such as a feed-forward neural network or an algorithm to calculate linear-prediction coefficients from time series, and given a prior distribution for the features, there exists a unique generative model for the input data, which subject to mild requirements, is maximum entropy (MaxEnt) among all probability density functions (PDFs) that are consistent with the given feature prior. In this paper, we consider the problem of sampling from these MaxEnt projected PDFs. The sampling process consists of drawing a sample from the given feature prior distribution, then drawing samples uniformly distributed on the inversion set (set of input samples consistent with the drawn feature value, usually a manifold). The process is called uniform manifold sampling (UMS). We describe UMS for simple nonlinear and iterative feature transformations, then focus on linear transformations with input data constraints (xi > 0 or 0≤ xi ≤ 1), which require MCMC-based sampling. We discover that the manifold centroid (sample mean for a fixed feature value) is useful as a deterministic MaxEnt feature inversion solution. We show how to predict the centroid efficiently without sampling and demonstrate its usefulness in speeding up MCMC by an order of magnitude, and in spectral estimation and image reconstruction. Finally, we provide an example of UMS in a classification experiment in which we use Monte Carlo integration to create true generative models from arbitrary classifiers. © 1991-2012 IEEE.


News Article | November 3, 2016
Site: phys.org

As the polar ice caps are melting, the Northwest Passage is becoming more navigable – its economic use is within reach. Even the federal government of Germany has the topic on the agenda. "For a country dependent on exports, such as Germany, shorter sea routes are of great importance," says Dr. Wolfgang Koch, Head of Sensor Data and Information Fusion at the Fraunhofer Institute for Communication, Information Processing and Ergonomics FKIE. "Compared to the route between East Asia and Europe via the Suez Canal, the route via the Northwest Passage is about 5,000 nautical miles shorter, which means enormous savings for ship owners," says Koch. This route has been too dangerous, though: There is no reliable information about it. That is what Koch and a German-Canadian research team want to change. In the project PASSAGES (Protection and Advanced Surveillance System for the Artic: Green, Efficient, Secure), they are conducting the preparatory work for safe navigation through the icy waters. This is not an easy task, because the route is challenging not only due to the presence of a number of bays, islands, uncharted shoals and narrows, but also because of drift ice and extreme weather conditions. In addition other vessels sometimes do not record their positions or (in cases like those of illegal fishing vessels) record false positions. The system has to provide reliable information about all of that. The problem: There is little data available, because there is no infrastructure in sensor and communication technology. The entire route is larger than Western Europe and is sparsely populated. Even if there were sufficient data these would have to be processed into information which would be useful for stakeholders such as ship crews. The researchers are addressing all of these difficulties by developing concepts about which technology can gather what information and where and how it can be fused. "The difficulty is to bring together very diverse and inaccurate data in order to provide for instance decision support for ship captains to decide on which route to take when," says Koch. Creating algorithms for the fusion of sensor data is the specialty of the FKIE researchers. The first thing to do, though, is to tap data sources in the harsh climate zone. For example, the Automatic Identification System (AIS) can be used, which reports among other information ships' current position. In addition, there are satellite images (which, however, have the drawback of being incomplete). Even old sonar systems from the Cold War era could be resurrected. However: A high-resolution view by means of which captains could safely pilot through the water does not yet exist. Koch is therefore still hoping for another data provider: the passive radar. This technology uses the electromagnetic pollution from mobile stations near the coast. Receiving stations will tap into this and gain information about ships and ice blocks – their size, position and speed. "In this way, large areas can be monitored," says Koch. Other possibilities are unmanned vehicles that gather information below and above water. Based on the newly gained insights and ideas for such a system, Koch would like to establish an appropriate monitoring and information system: "We hope that our research project will be followed by a development project." The project is supported by the German Federal Ministry for Economic Affairs and Energy (BMWi) as well as the project partners Airbus, exact Earth, and Dalhousie University in Halifax, Canada. An operational system would be a great asset for shipping companies, the Coast Guard and maritime authorities. Even insurance companies are interested in the data. "On this basis, they could calculate premiums for the ships that are to be insured," explains Koch. After the tricky route is made manageable through safe navigation, ship owners would not have to spend so much on insurance coverage. It will take at least a decade until the navigation system is ready to be launched. Over these years, the route will become progressively ice-free and economically navigable for shipping traffic. This will show that the scientific groundwork of today will have been worthwhile. Koch: "As dramatic as global warming is, we try to get at least something positive out of it." Explore further: Bringing the human element to ship design


News Article | November 3, 2016
Site: www.sciencedaily.com

The Northwest Passage is becoming navigable for longer periods of the year. Ship traffic, however, still bears hazardous risks. A German-Canadian research team co-initiated by Fraunhofer wants to change that. In the project PASSAGES, it is conducting the preparatory work for a safe navigation through the icy waters. As the polar ice caps are melting, the Northwest Passage is becoming more navigable -- its economic use is within reach. Even the federal government of Germany has the topic on the agenda. "For a country dependent on exports, such as Germany, shorter sea routes are of great importance," says Dr. Wolfgang Koch, Head of Sensor Data and Information Fusion at the Fraunhofer Institute for Communication, Information Processing and Ergonomics FKIE. "Compared to the route between East Asia and Europe via the Suez Canal, the route via the Northwest Passage is about 5,000 nautical miles shorter, which means enormous savings for ship owners," says Koch. This route has been too dangerous, though: There is no reliable information about it. That is what Koch and a German-Canadian research team want to change. In the project PASSAGES (Protection and Advanced Surveillance System for the Artic: Green, Efficient, Secure), they are conducting the preparatory work for safe navigation through the icy waters. This is not an easy task, because the route is challenging not only due to the presence of a number of bays, islands, uncharted shoals and narrows, but also because of drift ice and extreme weather conditions. In addition other vessels sometimes do not record their positions or (in cases like those of illegal fishing vessels) record false positions. The system has to provide reliable information about all of that. The problem: There is little data available, because there is no infrastructure in sensor and communication technology. The entire route is larger than Western Europe and is sparsely populated. Even if there were sufficient data these would have to be processed into information which would be useful for stakeholders such as ship crews. The researchers are addressing all of these difficulties by developing concepts about which technology can gather what information and where and how it can be fused. "The difficulty is to bring together very diverse and inaccurate data in order to provide for instance decision support for ship captains to decide on which route to take when," says Koch. Creating algorithms for the fusion of sensor data is the specialty of the FKIE researchers. The first thing to do, though, is to tap data sources in the harsh climate zone. For example, the Automatic Identification System (AIS) can be used, which reports among other information ships' current position. In addition, there are satellite images (which, however, have the drawback of being incomplete). Even old sonar systems from the Cold War era could be resurrected. However: A high-resolution view by means of which captains could safely pilot through the water does not yet exist. Koch is therefore still hoping for another data provider: the passive radar. This technology uses the electromagnetic pollution from mobile stations near the coast. Receiving stations will tap into this and gain information about ships and ice blocks -- their size, position and speed. "In this way, large areas can be monitored," says Koch. Other possibilities are unmanned vehicles that gather information below and above water. Based on the newly gained insights and ideas for such a system, Koch would like to establish an appropriate monitoring and information system: "We hope that our research project will be followed by a development project." The project is supported by the German Federal Ministry for Economic Affairs and Energy (BMWi) as well as the project partners Airbus, exact Earth, and Dalhousie University in Halifax, Canada. An operational system would be a great asset for shipping companies, the Coast Guard and maritime authorities. Even insurance companies are interested in the data. "On this basis, they could calculate premiums for the ships that are to be insured," explains Koch. After the tricky route is made manageable through safe navigation, ship owners would not have to spend so much on insurance coverage. It will take at least a decade until the navigation system is ready to be launched. Over these years, the route will become progressively ice-free and economically navigable for shipping traffic. This will show that the scientific groundwork of today will have been worthwhile. Koch: "As dramatic as global warming is, we try to get at least something positive out of it."


News Article | September 1, 2016
Site: phys.org

Ships are a safe means of transport. When something happens, human error is usually the cause – according to the latest studies, in 80 percent of the accidents. Economic and technical aspects are often the primary focus when ships are being planned and designed. How the crew can be included in the on-board procedures then either takes second chair or gets left out completely. "Established ergonomic concepts that improve the interaction between man and machine have, up to now, scarcely been implemented in the maritime industry. The reason often lies in the lack of communication or understanding between crews and engineers, who build the vessels and approve them. Even when there is a transfer of available knowledge, there are often still barriers to implementing it, such as extra costs and inherent risk in changes to a technically proven design," says Dr. Eric Holder of the Human-Machine System's Department of the Fraunhofer Institute for Communication, Information Processing and Ergonomics FKIE in Wachtberg near Bonn. The maritime industry at one table In the EU-sponsored CyClaDes (Crew-Centered Design and Operation of Ships and Ship Systems) project, the FKIE sat at one table with universities, professional organizations, shipbuilders, ship owners, government agencies and ship crews. The goal of the collaboration: To better integrate the "human factor" in the development phase and the life cycle of a ship – to make shipping even safer. The interdisciplinary team extensively analyzed the places on ships where human needs have been neglected up to now, discovered potential areas where processes could be improved and sent the latest research findings to the relevant parties involved. The project resulted in an e-learning platform, guidelines, checklists and a book for naval architects, which can be put to use immediately by the maritime industry. Among other things, in cooperation with the World Maritime University, the FKIE scientists defined, planned and conceived several modules for an e-learning platform: The platform provides training scenarios for each stakeholder group. Regardless of whether they were a crew member, ship owner, regulatory agency, systems designer or naval architect, each can find information and instructions here on the topic of user-friendly design. The book "Improving ship operational design", published by the Nautical Institute and written by CyClaDes members and experts in the field, provides naval architects with concrete requirements that crews face at sea. This knowledge is critical since these days architects and engineers seldom go to sea and sailors hardly have any opportunities to meet with them and exchange ideas. Government agencies can use the guidelines and checklists developed during CyClaDes in regulatory standards and design guidelines or directly in the ship design audit. "However, the industry only produces the concepts if they have practical uses," explains Holder. The FKIE examined the navigation bridge and the machine room more closely. "Both of these places play a decisive role in the safety on ships. If something goes wrong here, it can quickly become critical," emphasizes Holder. Communication here often takes place via signals or the ship's telephone. The perspectives from each department on the moving ship can be very different: the clear view from the bridge or an isolated room inside the ship; above the captain wants to move ahead as quickly as possible, below the engineers see the safety and durability of their engines first. The FKIE methodologically analyzed the communication processes and overlapping tasks between the two departmets and spoke in depth with all of those involved. Then, the scientists developed the concept for a touch-screen planning table, which demonstrates the potential of the user-oriented approach in concrete terms. In this concept the bridge and engineers come together in person, have all the information they need at one glance and can plan and discuss together how to proceed.


Feldmann M.,Fraunhofer Institute for Communication, Information Processing and Ergonomics | Franken D.,Airbus | Koch W.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
IEEE Transactions on Signal Processing | Year: 2011

The task of tracking extended objects or (partly) unresolvable group targets raises new challenges for both data association and track maintenance. Due to limited sensor resolution capabilities, group targets (i.e., a number of closely spaced targets moving in a coordinated fashion) may show a similar detection pattern as extended objects, namely a varying number of detections whose spread is determined by both the statistical sensor errors as well as the physical extension of the group or extended object. In both cases, tracking and data association under the one targetone detection assumption are no longer applicable. This paper deals with the problem of maintaining a track for an extended object or group target with varying number of detections. Herein, object extension is represented by a symmetric positive definite random matrix. A recently published Bayesian approach to tackling this problem is analyzed and discussed. From there, a new approach is derived that is expected to overcome some of the weaknesses the mentioned Bayesian approach suffers from in certain applications. © 2010 IEEE.


Wieneke M.,Fraunhofer Institute for Communication, Information Processing and Ergonomics | Koch W.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
IEEE Transactions on Aerospace and Electronic Systems | Year: 2012

Conventional tracking algorithms rely on the assumption that the targets of interest are point source objects. However, in realistic scenarios the point source assumption is often not suitable and estimating the object extent becomes a crucial aspect. Recently, a Bayesian approach to extended object tracking using random matrices has been proposed. Within this approach, ellipsoidal object extensions are modeled by random matrices and treated as additional state variables to be estimated. However, only a single-object solution has been presented so far. In this work we present the multi-object extent of this approach. We derive a new variant of probabilistic multi-hypothesis tracking (PMHT) that simultaneously estimates the ellipsoidal shape and the kinematics of each object using expectation-maximization (EM). Both the ellipsoids and the kinematic states are iteratively optimized by specific Kalman filter formulae that arise directly from the PMHT framework. The novel method is demonstrated and evaluated by simulations. © 2012 IEEE.


Govaers F.,Fraunhofer Institute for Communication, Information Processing and Ergonomics | Koch W.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
IEEE Transactions on Aerospace and Electronic Systems | Year: 2012

Track-to-track fusion (T2TF) aims at combining locally preprocessed information of individual trackers at a fusion center. Particularly, such schemes are obligatory in many applications of distributed sensors because of limited communication resources. If T2TF yields equivalent results compared with a Kalman filter (KF) processing all measurements from every sensor, it might be called optimal T2TF. It is well known that this can be achieved for deterministically moving targets, or if the local sensor tracks produced at all individual scan times are available in the fusion center. However, if such a full-rate communication is not available, achieving optimal performance is a great challenge due to cross-correlations between the local tracks. In this paper, we show that track decorrelation and therefore optimal T2TF can be achieved if sensor properties such as the measurement error covariance and measurement function are known at the each tracking system. To this end, local tracks have to be modified and local optimality has to be given up. As a result, we provide a distributed Kalman-type processing scheme for maneuvering targets, which yields optimal T2TF results at arbitrarily chosen instants of time by communicating and combining the local sensor tracks. Applications can be found in tracking scenarios with dynamically changing bandwidth constraints. © 2012 IEEE.


Koch W.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
IEEE Aerospace and Electronic Systems Magazine | Year: 2010

This tutorial paper provides a short introduction to selected aspects of sensor data fusion by discussing characteristic examples. We consider three cases when fusion of sensor data is important: when emphasis is placed on data produced at different instants of time (i.e., target tracking), when data being collected from different sensor sources are important, and when we have data with background information on the sensor performance as well as data. with nonsensor context information. The feedback from data processing to the data acquisition process is illustrated by a sensor management example. © 2010 IEEE.


Kaune R.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
15th International Conference on Information Fusion, FUSION 2012 | Year: 2012

In sensor networks, passive localization can be performed by exploiting the received signals of unknown emitters. In this paper, the Time of Arrival (TOA) measurements are investigated. Often, the unknown time of emission is eliminated by calculating the difference between two TOA measurements where Time Difference of Arrival (TDOA) measurements are obtained. In TOA processing, additionally, the unknown time of emission is to be estimated. Therefore, the target state is extended by the unknown time of emission. A comparison is performed investigating the attainable accuracies for localization based on TDOA and TOA measurements given by the Cramér-Rao Lower Bound (CRLB). Using the Maximum Likelihood estimator, some characteristic features of the cost functions are investigated indicating a better performance of the TOA approach. But counterintuitive, Monte Carlo simulations do not support this indication, but show the comparability of TDOA and TOA localization. © 2012 ISIF (Intl Society of Information Fusi).


Nickel U.R.O.,Fraunhofer Institute for Communication, Information Processing and Ergonomics
IEEE International Symposium on Phased Array Systems and Technology | Year: 2010

The increasing number of communication transmitters makes passive radar an attractive and simple alternative to active radar systems. GSM base stations as illuminators for passive radar enjoy the advantage of digital waveform, frequency diversity and global availability. The problems with these transmitters are the low transmit power and the more or less directive transmit patterns. These limit the achievable range. This paper studies how a useful system can be designed by fusion of bistatic systems using multiple transmitters. © 2010 IEEE.

Loading Fraunhofer Institute for Communication, Information Processing and Ergonomics collaborators
Loading Fraunhofer Institute for Communication, Information Processing and Ergonomics collaborators