"Index exceeds the number of array elements" in YoloV2ObjectDetector detect function.

2 次查看（过去 30 天）
Ryan Comeau
Ryan Comeau 2020-3-25
回答： Anurag Pratap Singh 2020-6-22
After running the following code, I complete the training for the YOLOV2 and obtain the detector object. When I try to use the detector, I obtain the following compile error:
THIS IS FOR RELEASE R2020a despite saying R2019b. R2020a is not a available option in drop down menu.
Index exceeds the number of array elements (2).
Error in yolov2ObjectDetector>iPredictUsingFeatureMap (line 868)
featureMap = reshape(featureMap,gridSize(1)*gridSize(2),gridSize(3),1,[]);
Error in yolov2ObjectDetector>iPostProcessActivations (line 982)
outputPrediction = iPredictUsingFeatureMap(featureMap, params.Threshold, info.PreprocessedImageSize,
anchorBoxes, params.FractionDownsampling, params.WH2HW);
Error in yolov2ObjectDetector>iPredictUsingDatastore (line 931)
iPostProcessActivations(fmap, batchInfo{ii}, anchorBoxes, params);
Error in yolov2ObjectDetector/detect (line 397)
varargout{1} = iPredictUsingDatastore(ds, this.Network, params, anchors, layerName);
Please help, thanks in advance.
%train set
imdsTrain = imageDatastore(table1{:,'imagefilename'},'ReadFcn',@fitsread);
bldsTrain = boxLabelDatastore(traintbl);
trainData = combine(imdsTrain, bldsTrain);
imdsTest = imageDatastore(table2{:,'imagefilename'},'ReadFcn',@fitsread);
bldsTest = boxLabelDatastore(testtbl);
testData = combine(imdsTest, bldsTest);
layers = [
imageInputLayer([2560 2560],"Name","imageinput")
convolution2dLayer([40 40],48,"Name","conv_1","Padding","same","Stride",[7 7])
batchNormalizationLayer("Name","batchnorm_1")
reluLayer("Name","relu_1")
maxPooling2dLayer([2 2],"Name","maxpool_1","Padding","same","Stride",[2 2])
convolution2dLayer([25 25],128,"Name","conv_2","Padding","same","Stride",[5 5])
batchNormalizationLayer("Name","batchnorm_2")
reluLayer("Name","relu_2")
convolution2dLayer([15 15],128,"Name","conv_4","Padding","same","Stride",[4 4])
batchNormalizationLayer("Name","batchnorm_4")
reluLayer("Name","relu_4")
maxPooling2dLayer([2 2],"Name",'maxpool_2',"Padding","same",'stride',[2 2])
convolution2dLayer([9 9],128,"Name","conv_6","Padding","same","Stride",[3 3])
batchNormalizationLayer("Name","batchnorm_6")
reluLayer("Name","relu_6")
convolution2dLayer([9 9],128,"Name","conv_5","Padding","same","Stride",[3 3])
batchNormalizationLayer("Name","batchnorm_5")
reluLayer("Name","relu_5")
maxPooling2dLayer([2 2],"Name","maxpool_3","Padding","same",'stride',[2 2])
convolution2dLayer([7 7],128,"Name","conv_9","Padding","same","Stride",[2 2])
batchNormalizationLayer("Name","batchnorm_9")
reluLayer("Name","relu_9")
convolution2dLayer([7 7],128,"Name","conv_8","Padding","same","Stride",[2 2])
batchNormalizationLayer("Name","batchnorm_8")
reluLayer("Name","relu_8")];
lgraph_homemade=layerGraph(layers);
%%%%%%%%%%make our own yolo from resnet50
numAnchors = 7;
[anchorBoxes,~] = estimateAnchorBoxes(trainData,numAnchors);
featureLayer = 'relu_8';
inputSize = [2560 2560];
numClasses = 1;
lgraph2 = yolov2Layers(inputSize,numClasses,anchorBoxes,lgraph_homemade,featureLayer);
options = trainingOptions('adam',...
'InitialLearnRate',0.005,...
'LearnRateSchedule','piecewise', ...
'LearnRateDropFactor',0.5, ...
'LearnRateDropPeriod',1, ...
'Verbose',true,...
'MiniBatchSize',8,...
'MaxEpochs',4,...
'Shuffle','never',...
'VerboseFrequency',1);
[detector,info] = trainYOLOv2ObjectDetector(trainData,lgraph2,options);
res=detect(detector,testData)

请先登录，再回答此问题。

采纳的回答

Anurag Pratap Singh
Hi Ryan
It is my understanding that you are trying to use the detector object.However you're getting the error Index exceeds the number of array elements (2).
This error is occuring because on line868 the gridSize array is of size 2(which is also indicated in the error) and you are accesing the 3rd element in gridSize by gridSize(3) which is not present.
Please refer to the https://www.mathworks.com/help/matlab/math/array-indexing.html (array indexing) documentation for more information on accessing the array elements.

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Recognition, Object Detection, and Semantic Segmentation 的更多信息

标签

产品


版本

R2019b

Community Treasure Hunt

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

Start Hunting!

Translated by