Device and method for decision support of an artificial cognitive system

Patent metadata

Field	Value
Publication number	US-12499670-B2
Application number	US-202117907221-A
Country	US
Kind code	B2
Filing date	Mar 8, 2021
Priority date	Mar 23, 2020
Publication date	Dec 16, 2025
Grant date	Dec 16, 2025

Abstract

A device for decision support of a cognitive system based on data originating from a plurality of data sources, the device including a processing unit associated with each data source, each processing unit including an encoding unit configured to determine, from the data from the data source associated with the processing unit, a representation of data in a common representation space by applying a machine learning algorithm to the data, the device further including a data fusion unit configured to determine a representation model of an environment of the cognitive system by combining the data representations determined by the encoding units associated with the plurality of data sources through the application of a data fusion algorithm.

First claim

The invention claimed is:

1 . A decision assistance device for a cognitive system based on data from a plurality of data sources, the device comprising:
a respective processing circuit associated with each data source, each respective processing circuit comprising an encoder configured to determine, based on data from the corresponding data source associated with the respective processing circuit, a corresponding data representation in a common representation space; and data fusion circuitry configured to determine a representation model of an environment of the cognitive system by combining the data representations determined by the corresponding encoders associated with the plurality of data sources by applying a data fusion algorithm, wherein the data fusion algorithm is a recurrent neural network type of machine-learning algorithm that synchronizes and temporally aligns the data from the plurality of data sources, wherein each respective processing circuit associated with the corresponding data source is configured to determine, based on the determined representation model of the environment, the corresponding reconstructed data representation, each encoder is configured to implement, during a training phase, a machine learning algorithm using training data, and a particular processing circuit is further configured to determine, based on the determined representation model of the environment, a particular reconstructed data representation associated with a missing data source associated with a particular encoder that is used during a preliminary training phase of training said device, but inhibited during a usage phase of using said device.

2 . The device as claimed in claim 1 , wherein the particular processing circuit uses an algorithm originating from machine learning that uses generative adversarial networks.

3 . The decision assistance device as claimed in claim 2 , wherein each respective processing circuit is further configured to compare the data from the corresponding data source associated with said processing circuit with the corresponding reconstructed data representation determined by the respective processing circuit associated with the corresponding data source.

4 . The decision assistance device as claimed in claim 3 , further comprising decision circuitry configured to determine an action to be implemented by the cognitive system based on said representation model of the environment and/or on said comparison.

5 . The device as claimed in claim 1 , wherein the particular processing circuit corresponds to a virtual sensor, the decision assistance device receiving images from a rear-left camera and a rear-right camera of a vehicle.

6 . The device as claimed in claim 1 , wherein the training data are data derived from the corresponding reconstructed data representation determined by the respective processing circuit.

7 . The decision assistance device as claimed in claim 1 , wherein said machine learning algorithm is a neural network.

8 . The decision assistance device as claimed in claim 7 , wherein said neural network is a sequential multi-channel auto-encoder.

Assignees

• Renault Sas
• Ampere Sas

Inventors

• Ancora Andrea
• Da-Silva-Filarder Matthieu
• Derome Maxime
• Filippone Maurizio
• Michiardi Pietro

Classifications

• G06V20/56

  exterior to a vehicle by using sensors mounted on the vehicle · CPC title

• G06V2201/07

  Target detection · CPC title

• G06V10/82

  using neural networks · CPC title

• G06N3/096

  Transfer learning · CPC title

• G06N3/094

  Adversarial learning · CPC title