Robert Boschgmbh

Christoph Gladisch

Christian Heinzemann

Thomas Heinz

Matthias Woehrle

Joachim Sohns

Daniel Seiler-Thull

Ji Su Yoon

Philipp Glaser

Jens Oehlerking

Martin Herrmann

Michael Rittel

Nadja Schalm

Andreas Heyl

Ulrich Seger

Konrad Groh

Oliver Willers

Martin Butz

Tino Brade

A computer-implemented method for evaluating an image classifier, in which a classifier output of the image classifier is provided for the actuation of an at least semi-autonomous robot. The evaluation method includes: ascertaining a first dataset including image data and annotations being assigned to the image data, the annotations including information about the scene imaged in the respective image and/or about image regions to be classified and/or about movement information of the robot; ascertaining regions of the scenes that are reachable by the robot based on the annotations; ascertaining relevance values for image regions to be classified by the image classifier; classifying the image data of the first image dataset with the aid of the image classifier; evaluating the image classifier based on image regions correctly classified by the image classifier and incorrectly classified image regions, as well as the calculated relevance values of the corresponding image regions.

Method and device for evaluating an image classifier

A computer-implemented method for analysing the relevance of visual parameters for training a computer vision model. Upon adjusting the set of visual parameters to increase their relevance, a new set of visual data and corresponding groundtruth results that can be used in (re)training and/or testing the computer vision model.

Computer-implemented method for analyzing relevance of visual parameters for training a computer vision model

A computer-implemented method for testing conformance between images generated by a synthetic image generator and images obtained from authentic visual data. A conformance test result results from comparing results of global sensitivity analyses used to rank the effect of visual parameters on the computer vision model both for synthetic and authentic visual data.

Computer-implemented method for testing conformance between real and synthetic images for machine learning

A method for controlling a vehicle. In the method, data of a digital road map are read in, zones are determined for the digital road map, and possible sequences of trips along a road of the digital road map are ascertained as a function of the determined zones. Furthermore, it is ascertained, as a function of sensor data and/or current driving data of the vehicle, whether a current or predicted traffic situation is outside the possible sequences or corresponds to a possible sequence that is determined as being outside an intended operating range. If the current or predicted traffic situation is outside the possible sequences or corresponds to the possible sequence outside the intended operating range, a measure is determined and the vehicle is controlled as a function of the measure that is taken.

Method for controlling a vehicle

Modifying a visual parameter specification characterising the operational design domain of the computer vision model by improving the visual parameter specification according to a sensitivity analysis of the computer vision model.

Modifying parameter sets characterising a computer vision model

Reducing the number of parameters in a visual parameter set based on a sensitivity analysis of how a given visual parameter affects the performance of a computer vision model to provide a verification parameter set having a reduced size.

Verification of computer vision models

A method for testing a technical system. Tests are carried out with the aid of a simulation of the system. The tests are evaluated with respect to a fulfillment measure of a quantitative requirement on the system and different error measures of the simulation. On the basis of the fulfillment measure and each of the error measures, a classification of the tests is carried out as either reliable or unreliable case by case. A selection among the error measures is made on the basis of a number of the tests classified as reliable.

Method and device for testing a technical system

A method for evaluating a simulation model. In the method, for selected test cases, a first performance index is calculated in the simulation model. For the same test case, a second performance index is ascertained in a real test environment. For each of the test cases, a difference is calculated between the first performance index and the second performance index, and a signal metric is determined. For each of the signal metrics, an interrelation between the difference and the respective signal metric is investigated. The signal metric that exhibits the closest interrelation with the difference is selected.

Method and apparatus for evaluating and selecting signal comparison metrics

A method or testing a system. Input parameters of the system are divided into a first group and a second group. Using a first method, a first selection is made from among the input parameter assignments of the first group. Using a second method, a second selection is made from among the input parameter assignments of the second group. A characteristic value is calculated from the second selection. The first selection is adapted depending on the characteristic value.

Method and apparatus for testing a system, for selecting real tests, and for testing systems with machine learning components

A method for testing a technical system. The method includes: tests are carried out with the aid of a simulation of the system, the tests are evaluated with respect to a fulfillment measure of a quantitative requirement on the system and an error measure of the simulation, on the basis of the fulfillment measure and error measure, a classification of the tests as either reliable or unreliable is carried out.

A method for testing a technical system. The method includes: tests are carried out with the aid of a simulation of the system, the tests are evaluated with respect to a fulfillment measure of a quantitative requirement on the system and an error measure of the simulation, on the basis of the fulfillment measure and error measure, a classification of the tests as either reliable or unreliable is carried out, and a test database is improved on the basis of the classification.

Growing community of inventors

Renningen, Germany

Christoph Gladisch