trainNetwork error Size of Matrix Training Data for CNN Regression

2 次查看（过去 30 天）

Hao Mai 2018-4-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/395388-trainnetwork-error-size-of-matrix-training-data-for-cnn-regression

编辑： Hao Mai 2018-4-17

Description: I wanna transfer a DnCnn Training network to denoise a new 2-D data But it always send Error: X and Y must have the same number of observations. I only use 1 data to train and reshape the data as a formal dataset but it still did not work Very Need you help

Code: %Load Data load('Synmodule.mat'); load('SynNoise.mat');

Ximds=TraceDataNorm(s,0,1); Yimds = TraceDataNorm(SynNoise,0,1)-Ximds;

%Load Pretrained Network

XNew = zeros(size(Ximds,1),size(Ximds,2),1,1); XNew(:,:,1,1) = Ximds(:,:);

YNew = zeros(size(Yimds,1),size(Yimds,2),1,1); YNew(:,:,1,1) = Yimds(:,:);

net = denoisingNetwork('dncnn'); layers = dnCNNLayers();

%Train Network maxEpochs = 30; initLearningRate = 0.1; l2reg = 0.0001; batchSize = 128;

options = trainingOptions('sgdm',... 'Momentum',0.9,... 'InitialLearnRate',initLearningRate,... 'LearnRateSchedule','piecewise',... 'L2Regularization',l2reg,... 'MiniBatchSize',batchSize,... 'MaxEpochs',maxEpochs,... 'Plots','training-progress');

netTransfer = trainNetwork(XNew,Yimds,layers,options);

Error: nnet.internal.cnn.util.NetworkDataValidator/assertXAndYHaveSameNumberOfObservations (line 142) Invalid training data. X and Y must have the same number of observations.