DeconvNetToolbox
trainAll.m File Reference

The entry point to train a generic deconvolutional network with any number of layers and any generic connections back down to the input image. More...

Go to the source code of this file.


Detailed Description

The entry point to train a generic deconvolutional network with any number of layers and any generic connections back down to the input image.

All the parameters of the model can be set here in a 'model' struct. Alternatively, these parameters can be saved to a file title 'gui_has_set_the_params.mat' either by hand or using a gui (not included). Every parameters including the paths to the images you want to train on and the save directory are included in model and must be modified at the top of this file. This uses pooled maps instead of feature maps.

Author:
Matthew Zeiler
Date:
Feb 9, 2011
Parameters (Must Set in Script):
  • model the main structure with all the parameters of the model must be set within this file. There are many fields to set. Read the comments within the file for more information.
(Click for related files):

The entry point to train a generic deconvolutional network with any number of layers and any generic connections back down to the input image.

Definition in file trainAll.m.

 All Files Functions Variables Typedefs Enumerations Enumerator Defines