Abstract: A computer-implemented method for training a machine learning system. The training includes: determining, by an encoder of the machine learning system and based on a training input signal, a first intermediate representation characterizing a mean of a latent distribution of a latent space and a second intermediate representation characterizing a variance and/or covariance of the latent distribution; determining, based on the first intermediate representation and the second intermediate representation, a plurality of sigma points with respect to the latent distribution; determining an output signal, wherein the output signal is determined by providing a randomly sampled sigma point of the plurality of sigma points to a decoder of the machine learning system; adapting the machine learning system based on a loss value, wherein the loss value characterizes a difference between the training input signal and the output signal.

Abstract: A method for providing an object message about an object, recognized in surroundings of a road user, in a communication network for communicating with other road users. The road user includes a sensor system for detecting the surroundings and an evaluation unit for evaluating sensor data generated by the sensor system and transferring object messages via the communication network. The method includes: receiving sensor data, generated by the sensor system, in the evaluation unit; recognizing at least one object in the surroundings of the road user based on the sensor data, a movement parameter and a further object parameter being ascertained; calculating an object transfer priority; determining, based on the object transfer priority, whether the recognized object is to be included in an object message; and, if so, generating the object message including the recognized object, and sending the object message via the wireless communication network.

Abstract: A method for operating a vehicle for highly automated driving. The method includes a step of reading sensor data that comprise trip data of the vehicle from at least one acceleration sensor, at least one position sensor, and a velocity sensor, infrastructure data of infrastructure elements in a predefined environment of the vehicle from at least one environmental sensor, and object data of recognized traffic objects in the predefined environment from the environmental sensor. An environmental model for behavior planning and maneuver planning of the vehicle within the predefined environment is determined. The environmental model is determined by simultaneous localization and mapping using the sensor data and a factor graph into which the sensor data are integrated as factors. The environmental model is output to an interface to a planning device for behavior planning and maneuver planning of the vehicle.

Abstract: A method for classifying a driving behavior of a road user in the environment of an ego vehicle. The method includes: receiving trajectory data of a driving trajectory of a road user arranged in the environment of an ego vehicle; ascertaining a latent-space representation of the driving trajectory of the road user in a latent space; ascertaining a distance of the latent-space representation of the driving trajectory of the road user to a latent-space representation of at least one normal trajectory in the latent space and classifying a driving behavior of the road user as a normal driving behavior if the distance in the latent space falls below a predetermined limit value; or classifying the driving behavior of the road user as an abnormal driving behavior if the distance in the latent space exceeds the predetermined limit value.

Abstract: A method for classifying a driving behavior of a road user in the environment of an ego vehicle. The method includes: receiving trajectory data of a driving trajectory of a road user arranged in the environment of an ego vehicle; ascertaining a latent-space representation of the driving trajectory of the road user in a latent space; ascertaining a distance of the latent-space representation of the driving trajectory of the road user to a latent-space representation of at least one normal trajectory in the latent space; and classifying a driving behavior of the road user as a normal driving behavior if the distance in the latent space falls below a predetermined limit value; or classifying the driving behavior of the road user as an abnormal driving behavior if the distance in the latent space exceeds the predetermined limit value.

Abstract: A computer-implemented method for predicting future developments of a traffic scene includes aggregating scene-specific information about a traffic scene, and using a pre-trained encoder network to transform the aggregated scene-specific information into parameters of a multivariate probability distribution of latent features. The method further includes selecting samples of the multivariate probability distribution of latent features determined by the parameters, and using a pre-trained decoder network to transform each of the selected samples into an output set. The samples are selected deterministically, such that each selected sample represents a separate region of the multivariate probability distribution of the latent features, and the multivariate probability distribution of latent features is sampled in a raster-like manner via the totality of the selected samples.

Abstract: A method for vehicle control. The method includes a step of reading in a camera signal, a step of executing a semantic segmentation, and a step of ascertaining a control signal. The camera signal represents an optically detected image of a roadway to be driven by a vehicle. In the step of executing, the semantic segmentation of the image represented by the camera signal is executed, to detect a free area ahead of the vehicle from the image as a drivable road section. In addition, a roadway signal representing the detected drivable road section is provided. In the step of ascertaining, the control signal for activating at least one vehicle component is ascertained, using the roadway signal.

Abstract: A method for determining agent trajectories in a multi-agent scenario includes capturing, for each agent, previous trajectories of the agents and a vicinity of the agent in a local reference frame of the agent; and coding, for each agent, the previous trajectories of the agents, captured in the local reference frame of the agent, into trajectory feature vectors and the vicinity of the agent, captured in the local reference frame of the agent, into vicinity feature vectors using an encoder neural network. The method further includes processing, for each agent, the trajectory feature vectors, depending on one another and depending on the vicinity feature vectors, into local-context feature vectors using an attention-based neural network; and processing the local-context feature vectors for all agents into a global-context feature vector for each agent using a common attention-based neural transformation network.

Abstract: A computer-implemented system for predicting future developments of a traffic scene is proposed, with which a high significance of the prediction can be achieved and the computational effort for the prediction can be limited. For this purpose, the system includes a perception level for aggregating scene-specific information of an input scene, a backbone network for generating a feature set of latent features based on the scene-specific information, a classifier evaluating a specified number of different modes for the future developments of the input scene based on the feature set, and for each mode, a prediction module for generating a prediction for the future development of the input scene, wherein at least one prediction module can optionally be activated.

Abstract: A method for training a computer-implemented system for predicting future developments of a traffic scene is proposed, the system comprising at least a perception level for aggregating scene-specific information of an input scene, a backbone network for generating a feature set of latent features based on the scene-specific information, a classifier network that evaluates a specified number of different modes for the future developments of the input scene based on the feature set, and for each mode, a prediction module for generating a prediction for the future development of the input scene. According to the disclosure, the backbone network is trained along with the classifier network by modifying the weights of the backbone network and/or the weights of the classifier network such that a deviation between the learning phase evaluation of the classifier network and a realistic evaluation of the different modes is reduced.

Abstract: A method for fusing state data via a control unit. State data of a first mobile unit and of an object ascertained via a sensor system of the first mobile unit are received. State data of an object ascertained via a sensor system of a second mobile unit and/or state data of the second mobile unit, transmitted via a communication link from the second mobile unit to the first mobile unit, are received. A node is created in a time-position diagram for each set of received state data of the first mobile unit, the second mobile unit, and the objects. A data optimization of the state data ascertained by the first mobile unit and/or by the second mobile unit is carried out. An optimization problem is created based on the optimized state data ascertained by the first mobile unit and the optimized state data received from the second mobile unit.

Abstract: A method for exchanging pieces of maneuver information between vehicles. A parameterizable third-party trajectory planner provided by a third-party vehicle and mapping future pieces of maneuver information of the third-party vehicle are parameterized and executed in an ego vehicle, using at least one time parameter, to obtain at least one future third-party trajectory of the third-party vehicle.

Abstract: The present invention relates to a method of cooperatively coordinating future driving maneuvers of a vehicle with fellow maneuvers of at least one fellow vehicle, wherein trajectories for the vehicle are rated with an effort value each, trajectories and fellow trajectories of the fellow vehicle are combined into tuples, the trajectory and the associated effort value of a collision-free tuple are selected as reference trajectory and reference effort value, trajectories with a lower effort value than the reference effort value are classified as demand trajectories, trajectories with higher effort value than the reference effort value are classified as alternative trajectories, and a data packet having a trajectory set consisting of the reference trajectory and the associated reference effort value as well as at least one trajectory from a group comprising the demand trajectories and the alternative trajectories as well as the respective effort values is transmitted to the fellow vehicle.

Abstract: A method for providing a maneuver message for coordinating a maneuver between a road user and at least one further road user in a communication network. The method includes: receiving the communication data and/or the sensor data in the evaluation unit; determining a possible trajectory of the road user based on the communication data and/or the sensor data, at least one trajectory parameter describing a property of the possible trajectory being ascertained; calculating a trajectory transfer priority from the trajectory parameter, the trajectory transfer priority representing a relevance of the at least one possible trajectory for the road user and/or the further road user; determining, based on the trajectory transfer priority, whether the at least one possible trajectory is to be included in a maneuver message; if so: generating the maneuver message including the at least one possible trajectory, and sending the maneuver message via the communication network.

Abstract: A method for transferring a message in a communications network for communication between a road user and at least one further road user. The road user and the further road user each include an evaluation unit for transferring messages via the communications network. The method includes: receiving a first message in the evaluation unit, the first message including message segments, each including a priority value; determining an instantaneous capacity utilization of the communications network; filtering message segments to be transferred out of the first message, based on the priority values and the instantaneous capacity utilization of the communications network; and generating a second message including the message segments to be transferred, and sending the second message via the communications network.

Abstract: A method for selecting a main object for an assistance function or automated driving function of a driver assistance system or driving system of a motor vehicle, the system containing object selection branches which include a first rule-based object selection branch for selecting a main object for the function and a second object selection branch, the second object selection branch including an artificial neural network for selecting a main object for the function, having the following steps: Aggregating of sensor data of the at least one sensor to form one or more object data records; Evaluating a novelty of a traffic situation characterized by the aggregated object data records in relation to training data of the artificial neural network; Switching between the object selection branches of the system, a rule-based object selection branch being used when the novelty of the traffic situation exceeds a threshold value.

Abstract: The present invention relates to a method of cooperatively coordinating future driving maneuvers of a vehicle with fellow maneuvers of at least one fellow vehicle, wherein a fellow data packet is received from the fellow vehicle, in which a fellow trajectory set of a fellow reference trajectory is contained, a trajectory from a trajectory set for the vehicle is selected as a reference trajectory for the vehicle using the fellow reference trajectory, wherein a trajectory which is collision-free towards the fellow reference trajectory is selected, the trajectories of the trajectory set are rated using limit trajectories, and at least one cooperation trajectory is selected from the trajectories of the trajectory set using the reference effort value, wherein a data packet containing the reference trajectory and the cooperation trajectory is transmitted to the fellow vehicle.

Abstract: A method is provided for correcting a positional probability distribution, at least two mobile systems each ascertaining a positional probability distribution through respective GNSS receivers, at least one mobile system ascertaining a distance to at least one second mobile system, the at least two mobile systems exchanging the ascertained positional probability distribution among themselves through a communication link, and by using the at least two ascertained positional probability distributions and the distance between the at least two mobile systems, an improvement of the positional probability distributions being calculated. Furthermore, a method for providing at least one correction term is provided.

Abstract: A method is provided for maneuver planning and implementation for a vehicle by a vehicle-internal control unit. The control unit includes a strategic planning level for carrying out route planning, a tactical planning level for providing lane-accurate trajectories to possible destination points, and an operative planning level for selecting a destination point and a negotiable trajectory to the selected destination point. The planning levels have a cascading design. When each planning level is executed, an information exchange with neighboring vehicles is carried out via a communication link in order to ascertain collisions. When a collision is ascertained in at least one planning level, a maneuver coordination is carried out between the vehicles via the communication link. Moreover, a control unit is provided.

Abstract: A method for vehicle control. The method includes a step of reading in a camera signal, a step of executing a semantic segmentation, and a step of ascertaining a control signal. The camera signal represents an optically detected image of a roadway to be driven by a vehicle. In the step of executing, the semantic segmentation of the image represented by the camera signal is executed, to detect a free area ahead of the vehicle from the image as a drivable road section. In addition, a roadway signal representing the detected drivable road section is provided. In the step of ascertaining, the control signal for activating at least one vehicle component is ascertained, using the roadway signal.