A method of providing surrounding information of a vehicle includes: detecting at least one sensor-based obstacle using sensor detection information acquired through a surroundings sensor, detecting at least one image-based obstacle using image information acquired through a plurality of cameras, detecting at least one matched obstacle from the at least one sensor-based obstacle or the at least one image-based obstacle, and correcting the position of the at least one matched obstacle using at least one of a sensor-based detection result or an image-based detection result with respect to the at least one matched obstacle.

A method for processing autonomous driving simulation data. The method includes: determining a type of a message transmitted between a simulation system and an auto driving system (ADS); determining a data acquisition mode based on the type of the message; obtaining a data stream transmitted between the simulation system and the ADS based on the data acquisition mode; and determining performance of the ADS based on the data stream.

A method and device for sensor data fusion for a vehicle as well as a computer program and a computer-readable storage medium are disclosed. At least one sensor device (S1) is associated with the vehicle (F), and in the method, fusion object data is provided representative of a fusion object (O.sub.F) detected in an environment of the vehicle (F); sensor object data is provided representative of a sensor object (O.sub.S) detected by the sensor device (S1) in the environment of the vehicle (F); indicator data is provided representative of an uncertainty in the determination of the sensor object data; reference point transformation candidates of the sensor object (O.sub.S) are determined depending on the indicator data; and an innovated fusion object is determined depending on the reference point transformation candidates.

A processor of a device for providing a route according to an embodiment of the present invention may, on the basis of entering a first lane-maintaining driving mode, cause a vehicle to drive so as not to deviate from a first lane, and on the basis of a preset condition being satisfied, may enter a second lane-maintaining driving mode in which the vehicle is caused to drive in the first lane and a second lane adjacent to the first lane.

In various examples, a multi-sensor fusion machine learning model – such as a deep neural network (DNN) – may be deployed to fuse data from a plurality of individual machine learning models. As such, the multi-sensor fusion network may use outputs from a plurality of machine learning models as input to generate a fused output that represents data from fields of view or sensory fields of each of the sensors supplying the machine learning models, while accounting for learned associations between boundary or overlap regions of the various fields of view of the source sensors. In this way, the fused output may be less likely to include duplicate, inaccurate, or noisy data with respect to objects or features in the environment, as the fusion network may be trained to account for multiple instances of a same object appearing in different input representations.

Provided are an automatic parking control method and apparatus, relating to the industrial field of vehicles. The method comprises: when a speed of a vehicle is in a preset speed range and a time that the speed of the vehicle is in the preset speed range is greater than or equal to a preset time, collecting image data and radar data around the vehicle; inputting the image data and the radar data into a preset convolutional neural network model, and outputting at least one parking slot information; selecting target parking slot information according to a received parking-in selection operation; and according to the target parking slot information, generating a vehicle parking-in track for the vehicle to automatically park according to the vehicle parking-in track.

At a first timestep, one or more first objects can be determined in a first fusion image based on determining one or more first radar clusters in first radar data and determining one or more first two-dimensional bounding boxes in first camera data. First detected objects and first undetected objects can be determined by inputting the first objects and the first radar clusters into a data association algorithm, which determines first probabilities and adds the first radar clusters and the first objects to one or more of first detected objects or first undetected objects by determining a cost function. The first detected objects and the first undetected objects can be input to a first Poisson multi-Bernoulli mixture (PMBM) filter to determine second detected objects, second undetected objects and second probabilities. The second detected objects and the second undetected objects can be reduced based on the second probabilities determined by the first PMBM filter and the second detected objects can be output.

Systems and methods for controlling the operation of an autonomous vehicle are disclosed herein. One embodiment performs traffic light detection at an intersection using a sensor-based traffic light detector to produce a sensor-based detection output, the sensor-based detection output having an associated first confidence level; performs traffic light detection at the intersection using a vehicle-to-infrastructure-based (V2I-based) traffic light detector to produce a V2I-based detection output, the V2I-based detection output having an associated second confidence level; performs one of (1) selecting as a final traffic-light-detection output whichever of the sensor-based detection output and the V2I-based detection output has a higher associated confidence level and (2) generating the final traffic-light-detection output by fusing the sensor-based detection output and the V2I-based detection output using a first learning-based classifier; and controls the operation of the autonomous vehicle based, at least in part, on the final traffic-light-detection output.

A method in accordance with the invention controls movements of an actor, such as an autonomous vehicle. The method includes collecting one or more data sets of location information from sources that include a first actor, a second actor, and a fixed agent. Each data set indicates a current location of an object from one or more objects. The method also includes merging the data sets to generate a second set of current location information of the one or more objects, filtering the second set of current location information to remove location information of static objects to generate a subset of the current locations corresponding to a subset of the objects, using the subset of current locations to predict future locations of the subset of objects, and coordinating the movement of the first actor based on the predicted future locations to optimize one or more pre-determined criteria.

A method for calibrating a yaw rate sensor of a vehicle, comprises detecting a yaw rate of the vehicle from measurement data from the yaw rate sensor. A change in yaw angle is ascertained from sensor data from at least one optical surroundings sensor unit, wherein an offset of the yaw rate sensor is ascertained, the offset being ascertained by fusion of the detected yaw rate and the ascertained change in yaw angle. The yaw rate sensor is calibrated according to the ascertained offset.