MATLAB Answers

Creating a data store for CNN Training

3 views (last 30 days)
Thyagharajan K K
Thyagharajan K K on 9 May 2020
I am having a 4D array xy_RGBO(row,col,channel,sample)
This array has 1x11 values for each channel and four such channels are arranged in the 3rd dimension. This 3D block represents one image. The fourth dimension indicateds the index for each image. How can I create a data store to train CNN?
When I stored each image as TIFF file and the data store was created using
imds=imageDatastore(fullfile(rootFolder),'FileExtensions', '.tiff','IncludeSubfolders',true,'LabelSource','foldernames')
The following error appeared
Error in LeafDAGMultinput (line 160)
kkt_net = trainNetwork(imdsTrain,layers,opts); %data, layers, global training options
Caused by:
Layer 'input_stream_1': Missing input. Each layer input must be connected to the output of another layer.
Detected missing inputs:
input 'in2'
input 'in3'
Layer 'FCs_Connected': Missing input. Each layer input must be connected to the output of another layer.
Detected missing inputs:
input 'in2'

  0 Comments

Sign in to comment.

Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!