Simulated degraded sensor data may be generated for use in training a model. For instance, first sensor data collected by a sensor of a perception system of an autonomous vehicle may be received and converted into the simulated degraded sensor data for a particular degrading condition, such as a weather-related degrading condition. Then, the simulated degraded sensor data may be used to train a model for evaluating performance of the perception system to detect objects external to the autonomous vehicle under one or more conditions.

A driver assistance system and method are disclosed. The driver assistance system includes a first sensor installed at a vehicle and configured to have a field of view directed forward from the vehicle to acquire front image data, a second sensor selected from a group of radar and LIDAR sensors, installed at the vehicle, and configured to have a field of view directed forward from the vehicle to acquire front detection data, and a controller having a processor configured to process the front image data and the front detection data, wherein the controller is configured to detect a lane, in which the vehicle is traveling, or detect a front object located in front of the vehicle, in response to the processing of the image data and the front detection data, output a braking signal to a braking system of the vehicle when a collision between the vehicle and the front object is expected, and output a steering signal to a steering system of the vehicle when a collision between the vehicle and the front object is expected even with braking control.

A method may include obtaining first sensor data from a first sensor system and second sensor data from a second sensor system. The first and the second sensor systems may capture sensor data from a total measurable world. The method may include identifying a first object included in the first sensor data and a second object included in the second sensor data and determining first parameters corresponding to the first object and second parameters corresponding to the second object. The first parameters may be compared with the second parameters and whether the first object and the second object are a same object may be determined based on the comparing the first parameters and the second parameters. Responsive to determining that the first object and the second object are the same object, a set of objects representative of objects in the total measurable world including the same object may be generated.

Techniques for accurately predicting and avoiding collisions with objects detected in an environment of a vehicle are discussed herein. A vehicle safety system can implement a model to output data indicating an intersection probability between the object and a portion of the vehicle in the future. The model may employ a rear collision filter, a distance filter, and a time to stop filter to determine whether a predicted collision may be a false positive, in which case the techniques may include refraining from reporting such predicted collision to other another vehicle computing device to control the vehicle.

A sensor system for a vehicle includes a central module and a plurality of sub modules mounted in a frame of the vehicle, the sub modules being independently removable. The sub modules include sensors configured to capture image data and distance data in a vicinity of the vehicle. The central module is connected to each of the plurality of sub modules through a first network including a switching hub. The sub modules are individually connected to an external processor through a second network. The central processor is configured to synchronize the sub modules based on absolute time information through the first network, and the sub modules are configured to output the captured image data and distance data appended with synchronized time information to the external processor by communicating through the second network.

A method is described for determining the position of a vehicle equipped with a radar system that includes at least one radar sensor adapted to receive radar signals emitted from at least one radar emitter of the radar system and reflected the radar sensor. The method comprises: acquiring at least one radar scan comprising a plurality of radar detection points, wherein each radar detection point is evaluated from a radar signal received at the radar sensor and representing a location in the vicinity of the vehicle; determining, from a database, a predefined map, wherein the map comprises at least one element representing a static landmark in the vicinity of the vehicle; matching at least a subset of the plurality of radar detection points of the at least one scan and the at least one element of the map; deter-mining the position of the vehicle based on the matching.

Among other things, we describe techniques for detecting objects in the environment surrounding a vehicle. A computer system is configured to receive a set of measurements from a sensor of a vehicle. The set of measurements includes a plurality of data points that represent a plurality of objects in a 3D space surrounding the vehicle. The system divides the 3D space into a plurality of pillars. The system then assigns each data point of the plurality of data points to a pillar in the plurality of pillars. The system generates a pseudo-image based on the plurality of pillars. The pseudo-image includes, for each pillar of the plurality of pillars, a corresponding feature representation of data points assigned to the pillar. The system detects the plurality of objects based on an analysis of the pseudo-image. The system then operates the vehicle based upon the detecting of the objects.

A vehicle control system may include a controller that detects an object outside a vehicle, calculates an angle based on a ratio of a relative speed between the object and the vehicle to a speed of the vehicle, and updates a phase curve reflecting a phase distortion of an input signal based on the calculated angle.

An in-vehicle radar signal control method includes: determining a target interference area of a first vehicle, a vehicle in the target interference area interfering with an in-vehicle radar signal of the first vehicle; determining vehicles in the target interference area as a first vehicle cluster, and determining strength of in-vehicle radar signals of vehicles in the first vehicle cluster; determining whether a new second vehicle enters the target interference area; and in response to a determination that the second vehicle enters the target interference area, obtaining an adjustment signal; the adjustment signal indicating one or more of: increasing or reducing strength of the in-vehicle radar signal of the first vehicle, adjusting a travel speed of the first vehicle, and adjusting a travel direction of the first vehicle.

Techniques described herein are directed to comparing, using a machine-trained model, neural network activations associated with data representing a simulated environment and activations associated with data representing real environment to determine whether the simulated environment is causes similar responses by the neural network, e.g., a detector. If the simulated environment and the real environment do not activate the same way (e.g., the variation between neural network activations of real and simulated data meets or exceeds a threshold), techniques described herein are directed to modifying parameters of the simulated environment to generate a modified simulated environment that more closely resembles the real environment.