Computer-implemented method (100) for training a neural network (60) wherein the neural network (60) is configured to determine an output signal (y) based on an input signal (x) and wherein training comprises training parameters of a depth-wise convolutional layer of the neural network (60), wherein the depth-wise convolutional layer is initialized based on values drawn from a predefined probability distribution, wherein a variance of the probability distribution is characterized by a reciprocal of a square root of a number of filters applied at each depth of an input of the depth-wise convolutional layer.


    Access

    Download


    Export, share and cite



    Title :

    METHOD FOR INITIALIZING A NEURAL NETWORK


    Additional title:

    VERFAHREN ZUR INITIALISIERUNG EINES NEURONALEN NETZES
    PROCÉDÉ D'INITIALISATION D'UN RÉSEAU DE NEURONES


    Contributors:

    Publication date :

    2023-10-11


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    English


    Classification :

    IPC:    G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion / G06V



    Method for initializing neural network

    SCHMID FRANZISKA / LONG TAO | European Patent Office | 2024

    Free access

    METHOD FOR INITIALIZING A NEURAL NETWORK

    SCHMIDT FRANK / SACHSE TORSTEN | European Patent Office | 2023

    Free access

    METHOD FOR INITIALIZING A NEURAL NETWORK

    SCHMIDT FRANK / LONG TORSTEN | European Patent Office | 2023

    Free access

    Initializing Snakes

    Neuenschwander, W. / Fuo, P. / Szekely, G. et al. | British Library Conference Proceedings | 1994


    Initializing snakes [object delineation]

    Neuenschwander / Fua / Szekely et al. | IEEE | 1994