公开(公告)号：US20180075345A1

公开(公告)日：2018-03-15

申请号：US15264316

申请日：2016-09-13

Applicant: SAP SE

Inventor： Frank Gottfried ,  Bjoem Deiseroth ,  Burkhard Neidecker-Lutz

IPC: G06N3/063 ,  G06N3/04

Abstract: Technologies are provided for implementing temporal and spatio-temporal spiking neural networks (SNNs) using neuromorphic hardware devices. Temporal synapse circuits, with associated weight values, can be used to control spike times of connected neuron circuits. The controlled spike times of multiple neuron circuits can be used to temporally encode information in a neural network in neuromorphic hardware. Neuron circuits in a state space detection layer can be organized into multiple subsets. Neuron circuits in different subsets can be connected to output neuron circuits in an output layer by separate temporal synapse circuits. Spiking signals sent from the neuron circuits in the state space detection layer via separate temporal synapse circuits can cause associated output neuron circuits to generate output spiking signals at different times. The various spike times of the output neuron circuits can be aggregated to produce an output signal for the network.

公开(公告)号：US10671912B2

公开(公告)日：2020-06-02

申请号：US15264316

申请日：2016-09-13

Applicant: SAP SE

Inventor： Frank Gottfried ,  Bjoern Deiseroth ,  Burkhard Neidecker-Lutz

IPC: G06N3/063 ,  G06N3/04 ,  G06N20/00 ,  G06N3/08

Abstract: Technologies are provided for implementing temporal and spatio-temporal spiking neural networks (SNNs) using neuromorphic hardware devices. Temporal synapse circuits, with associated weight values, can be used to control spike times of connected neuron circuits. The controlled spike times of multiple neuron circuits can be used to temporally encode information in a neural network in neuromorphic hardware. Neuron circuits in a state space detection layer can be organized into multiple subsets. Neuron circuits in different subsets can be connected to output neuron circuits in an output layer by separate temporal synapse circuits. Spiking signals sent from the neuron circuits in the state space detection layer via separate temporal synapse circuits can cause associated output neuron circuits to generate output spiking signals at different times. The various spike times of the output neuron circuits can be aggregated to produce an output signal for the network.

公开(公告)号：US12147468B2

公开(公告)日：2024-11-19

申请号：US16219320

申请日：2018-12-13

Applicant: SAP SE

Inventor： Bjoern Deiseroth ,  Frank Gottfried

IPC: G06F16/51 ,  G06F16/532 ,  G06F16/587 ,  G06V10/34 ,  G06V10/46

Abstract: A method, a system, and a computer program product for performing on-demand feature extraction from a raw image of an object for analysis. A query is executed to retrieve an image of an object. Using one or more parameters of the query, a raw image of the object is compressed to generate a compressed image of the object. One or more features associated with the object are extracted from the compressed image of the object. Based on the compressed image of the object, the image of the object is generated using the extracted one or more features of the object.