Filter löschen
Filter löschen

DAG Network loading as Struct

11 Ansichten (letzte 30 Tage)
Rachel Ball
Rachel Ball am 6 Dez. 2022
Beantwortet: Rachel Ball am 16 Dez. 2022
I am trying to use a nerual network on a set of data. This is my code:
%This code is to create data to analyse in another program to test the performance of the NN.
clear;
clc;
%Open files
testfiledir = 'Z:\eee_biophotonics1\Shared\SPOT dataset (Training data)\Testing Images\123 images on 456 network\OCT';
matfiles = dir(fullfile(testfiledir, '*.tif'));
nfiles = length(matfiles); %finds length of file
data = cell(nfiles);
for i = 1 : nfiles
fid = fopen( fullfile(testfiledir, matfiles(i).name) ); %retrieves specific file
net = load('net_456.mat'); % This loads the network back into MATLAB
result = semanticseg(fid,net); % This segments an image "fid" using the network
filename='Z:\eee_biophotonics1\Shared\SPOT dataset (Training data)\Testing Images\123 images on 456 network\nn segmented';
save(result,'%d',i,'-tif');
fclose(fid);
end
However this error comes up:
Error using semanticseg
Expected net to be one of these types:
SeriesNetwork, DAGNetwork, dlnetwork
Instead its type was struct.
Error in semanticseg>iCheckNetwork (line 723)
validateattributes(net, {'SeriesNetwork', 'DAGNetwork', 'dlnetwork'}, ...
Error in semanticseg>iParseInputs (line 416)
iCheckNetwork(net);
Error in semanticseg (line 244)
params = iParseInputs(I, net, varargin{:});
Error in SPOTimageconversion_123_456 (line 12)
result = semanticseg(fid,net); % This segments an image "fid" using the network

Akzeptierte Antwort

Rachel Ball
Rachel Ball am 16 Dez. 2022
remove: net = load('net_456.mat')
In the current folder pane in matlab, select the network you want and click on it, this should run load("x.mat") in the console
It will then show you the name of the DAG network insde details pane, Put this name in the semanticseg function

Weitere Antworten (0)

Community Treasure Hunt

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

Start Hunting!

Translated by