Robert Boschgmbh

Sap Se

Michael Herman

Max Welling

Benedikt Alt

Herke Van Hoof

Elise Van Der Pol

Frans Adriaan Oliehoek

Volker Fischer

Matthias Kuehnel

Christoph Zimmer

Adrian Trachte

Michael Hilsch

Wenqing Liu

Joerg Wagner

Carolina Passenberg

Jan Peters

Dusan Radovic

Daniel Worrall

Eric Sebastian Schmidt

Hong Linh Thai

Jan Martin Lubisch

Thomas Kipf

A method for configuring a neural network which is designed to map measured data to one or more output variables. The method includes: transformation(s) of the measured data is/are specified which when applied to the measured data, is/are meant to induce the output variables supplied by the neural network to exhibit an invariant or equivariant behavior; at least one equation is set up which links a condition that the desired invariance or equivariance be given with the architecture of the neural network; by solving the at least one equation a feature is obtained that characterizes the desired architecture and/or a distribution of weights of the neural network in at least one location of this architecture; a neural network is configured in such a way that its architecture and/or its distribution of weights in at least one location of this architecture has/have all of the features ascertained in this way.

Configuring a neural network for equivariant or invariant behavior

Some embodiments are directed to a computer-implemented method of interacting with a physical environment according to a policy. The policy determines multiple action probabilities of respective actions based on an observable state of the physical environment. The policy includes a neural network parameterized by a set of parameters. The neural network determines the action probabilities by determining a final layer input from an observable state and applying a final layer of the neural network to the final layer input. The final layer is applied by applying a linear combination of a set of equivariant base weight matrices to the final layer input. The base weight matrices are equivariant in the sense that, for a set of multiple predefined transformations of the final layer input, each transformation causes a corresponding predefined action permutation of the base weight matrix output for the final layer input.

Physical environment interaction with an equivariant policy

A computer-implemented method for applying control to a robot, and apparatus therefor. A parametric model of an environment, in particular a deep neural network, is trained in accordance with a method for training the parametric model of the environment. The model is trained depending on a controlled system. A strategy is learned in accordance with a method for model-based learning of the strategy. Control is applied to the robot depending on the parametric model and on the strategy.

Method and device for controlling a robot

A control system for a motor vehicle, for outputting a controlled variable, with the aid of which a directly controlled variable of a motor vehicle is adjustable via suitable control operations, in order to adapt the directly controlled variable to a reference variable of the control system. The control system includes a controller, which is configured to output a first output variable on the basis of the directly controlled variable of the motor vehicle, and on the basis of the reference variable of the control system. The control system further includes a predictive model, which may be trained to output a second output variable that reflects a deviation of a driving behavior of a driver of the motor vehicle from the first output variable of the controller. The controlled variable of the control system encompasses an addition of the first output variable and the second output variable.

Control system for a motor vehicle and method for adapting the control system

A computer-implemented method for using machine learning to determine control parameters for a control system, in particular of a motor vehicle, in particular for controlling a driving operation of the motor vehicle. The method includes: providing a set of travel trajectories; deriving reward functions from the travel trajectories, using an inverse reinforcement learning method; deriving driver type-specific clusters based on the reward functions; determining control parameters for a particular driver type-specific cluster.

Method for determining control parameters for a control system

A computer-implemented method for inferring a device feature of a device produced on a wafer. The method includes: providing a wafer feature model associating a wafer position indicating a position of a produced device on the wafer to a device feature, wherein the wafer feature model is configured to be trained by one or more wafer feature maps and particularly configured as a Gaussian process model, providing a sample device feature of at least one device at a sample wafer position, and inferring the device feature of at least one other device of the wafer depending on the provided wafer feature model.

Method and device for determining a feature for devices produced on a wafer

A method for detecting an object with the aid of a system, the system having a plurality of modules connected in series one after the other, at least one connection between a predefinable module and its immediately following module being implemented in each case with the aid of a filter and storage module connected in series, the first module of the modules connected in series being connected to an acquisition unit, and the last module of the modules connected in series being connected to a detection module. The method includes temporal filtering.

Method, device, and computer program for detecting an object

Growing community of inventors

Sindelfingen, Germany

Michael Herman