Error using trainNetwork for image classification

Question

Sibi Isac 2018-1-5

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/375720-error-using-trainnetwork-for-image-classification

评论： Joss Knight 2018-1-9

I have following error from a classification.

Error using parallel.gpu.CUDADevice.hBuild An unexpected error occurred trying to retrieve CUDA device properties. The CUDA error was: CUDA_ERROR_UNKNOWN

Error in parallel.gpu.GPUDevice.getDevice (line 76) d = parallel.gpu.CUDADevice.hBuild( idx );

Error in parallel.gpu.GPUDevice.isAvailable (line 146) device = parallel.gpu.GPUDevice.getDevice( index );

Error in nnet.internal.cnn.util.isGPUCompatible (line 9) if(iCanUsePCT() && parallel.gpu.GPUDevice.isAvailable())

Error in nnet.internal.cnn.util.GPUShouldBeUsed (line 17) tf = nnet.internal.cnn.util.isGPUCompatible();

Error in trainNetwork>iSetupExecutionEnvironment (line 357) GPUShouldBeUsed = nnet.internal.cnn.util.GPUShouldBeUsed( ...

Error in trainNetwork (line 77) executionSettings = iSetupExecutionEnvironment( opts );

Error in imageClassification (line 34) myNet = trainNetwork(trainingImages, layers, opts); Here is the code i was trying, I couldn't figure out the reason behind the error.

clear;
close all;
clc;
% Import Training Data
trainFolder = 'Train';
testFolder = 'C:\Users\user\Google Drive\Sibi\Project\Data\test';
trainingImages = imageDatastore(trainFolder,'IncludeSubfolders',true,'FileExtensions',...
            {'.jpg','.tif'} ,'LabelSource','foldernames');
testImages = imageDatastore(testFolder,'FileExtensions',{'.jpg','.tif'});
% Prepare Training Sets
% minSetCount = min([trainingImages.Count]);
% trainingSets = partition(imageSets, minSetCount, 'randomize');
% [trainingImages, testImages] = splitEachLabel(images, 0.8, 'randomize');
net = alexnet;
layers = net.Layers;
layers(23) = fullyConnectedLayer(10); % change this based on # of classes
layers(25) = classificationLayer;
miniBatchSize = 64;
numIterationsPerEpoch = floor(numel(trainingImages.Labels)/miniBatchSize);
% functions = { ...
%     @plotTrainingAccuracy, ...
%     @(info) stopTrainingAtThreshold(info,95)};
opts = trainingOptions('sgdm',...
  'verbose',true,...
  'verboseFrequency', 1,...
  'InitialLearnRate', 0.001,...
  'OutputFcn',@plotTrainingAccuracy);
trainingImages.ReadFcn = @readImageFunction;
myNet = trainNetwork(trainingImages, layers, opts);
print -djpeg 'TrainingProgress.jpg';
testImages.ReadFcn = @readImageFunction;
save myNet myNet;
matobj = matfile('myNet.mat');%load the pre-trained net 
predictedLabels = classify(matobj.myNet, testImages); 
accuracy = mean(predictedLabels == testImages.Labels);
save accuracy accuracy;
function plotTrainingAccuracy(info)
persistent plotObj
if info.State == "start"
  plotObj = animatedline;
  xlabel("Iteration")
  ylabel("Training Accuracy")
elseif info.State == "iteration"
  addpoints(plotObj,info.Iteration,info.TrainingAccuracy)
  drawnow limitrate nocallbacks
end

end

I am using 2016a version of Matlab, can anyone help me sort out this problem?

>>

3 个评论
显示 1更早的评论隐藏 1更早的评论

Sibi Isac 2018-1-5

can i run this on the cpu?

Joss Knight 2018-1-9

Yes, just add 'ExecutionEnvironment', 'cpu' to the trainingOptions.

请先登录，再进行评论。

请先登录，再回答此问题。

Error using trainNetwork for image classification

3 个评论
显示 1更早的评论隐藏 1更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

Error using trainNetwork for image classification

3 个评论 显示 1更早的评论隐藏 1更早的评论

回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

3 个评论
显示 1更早的评论隐藏 1更早的评论