TRAINING A NEURAL NETWORK FOR IMAGE PROCESSING

TRAINING A NEURAL NETWORK FOR IMAGE PROCESSING

A mechanism for training a U-net neural network for image processing. During training, a U-net neural network is supplemented with an additional decoder. The additional decoder has fewer skip connections with the encoder of the U-net neural network than the original decoder of the U-net neural netwo...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: GOSHEN, Liran, GRINGAUZ, Asher
Format: Patent
Sprache:eng ; fre ; ger
Schlagworte:
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
PHYSICS
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:A mechanism for training a U-net neural network for image processing. During training, a U-net neural network is supplemented with an additional decoder. The additional decoder has fewer skip connections with the encoder of the U-net neural network than the original decoder of the U-net neural network, e.g., zero skip connections. The data produced by the original decoder and the additional decoder is used to define the value for a cost function used to modify the neural network during the training procedure.