Invalid argument name classificationmode name must be targetcategories, mask etc.

While using this code from Complex yolo document in R2023a version

doTraining = true;

if doTraining
    iteration = 0;
   
    % Create subplots for the learning rate and mini-batch loss.
    fig = figure;
    [lossPlotter, learningRatePlotter] = configureTrainingProgressPlotter(fig);

    % Custom training loop.
    for epoch = 1:maxEpochs
          
        reset(mbqTrain);
        shuffle(mbqTrain);
        
        while(hasdata(mbqTrain))
            iteration = iteration + 1;
           
            [XTrain,YTrain] = next(mbqTrain);
            
            % Evaluate the model gradients and loss using dlfeval and the
            % modelGradients function.
            [gradients,state,lossInfo] = dlfeval(@modelGradients,net,XTrain,YTrain,anchorBoxes,penaltyThreshold,networkOutputs);
    
            % Apply L2 regularization.
            gradients = dlupdate(@(g,w) g + l2Regularization*w, gradients, net.Learnables);
    
            % Determine the current learning rate value.
            currentLR = piecewiseLearningRateWithWarmup(iteration,epoch,learningRate,warmupPeriod,maxEpochs);
            
            % Update the network learnable parameters using the SGDM optimizer.
            [net,velocity] = sgdmupdate(net,gradients,velocity,currentLR);
    
            % Update the state parameters of dlnetwork.
            net.State = state;
            
            % Display progress.
            if mod(iteration,10)==1
                displayLossInfo(epoch,iteration,currentLR,lossInfo);
            end
                
            % Update training plot with new points.
            updatePlots(lossPlotter,learningRatePlotter,iteration,currentLR,lossInfo.totalLoss);
        end
    end
else
    net = mdl.net;
    anchorBoxes = mdl.anchorBoxes;
end





% Create a table to hold the bounding boxes, scores, and labels returned by
% the detector. 
results = table('Size',[0 3], ...
    'VariableTypes',{'cell','cell','cell'}, ...
    'VariableNames',{'Boxes','Scores','Labels'});

% Run the detector on images in the test set and collect the results.
reset(testData)
while hasdata(testData)
    % Read the datastore and get the image.
    data = read(testData);
    image = data{1,1};
    
    % Run the detector.
    executionEnvironment = 'auto';
    [bboxes,scores,labels] = detectComplexYOLOv4(net,image,anchorBoxes,classNames,executionEnvironment);
    
    % Collect the results.
    tbl = table({bboxes},{scores},{labels},'VariableNames',{'Boxes','Scores','Labels'});
    results = [results; tbl];
end

% Evaluate the object detector using the average precision metric.
metrics = evaluateDetectionAOS(results, testData)

Got this error

Error using dlarray/crossentropy

Invalid argument name 'ClassificationMode'. Name must be 'TargetCategories', 'Mask', 'Reduction', 'NormalizationFactor', 'DataFormat', or 'WeightsFormat'.

Error in complexYolotrial>@(a,b,c)crossentropy(a.*c,b.*c,'ClassificationMode','multilabel') (line 400)

objLoss = sum(cellfun(@(a,b,c) crossentropy(a.*c,b.*c,'ClassificationMode','multilabel'),objectnessPredCell,objectnessDeltaTarget,boxMaskTarget(:,2)));

Error in complexYolotrial>objectnessLoss (line 400)

objLoss = sum(cellfun(@(a,b,c) crossentropy(a.*c,b.*c,'ClassificationMode','multilabel'),objectnessPredCell,objectnessDeltaTarget,boxMaskTarget(:,2)));

Error in complexYolotrial>modelGradients (line 339)

objLoss = objectnessLoss(YPredCell(:,1),objectnessTarget,objectMaskTarget);

Error in deep.internal.dlfeval (line 17)

[varargout{1:nargout}] = fun(x{:});

Error in deep.internal.dlfevalWithNestingCheck (line 15)

[varargout{1:nargout}] = deep.internal.dlfeval(fun,varargin{:});

Error in dlfeval (line 31)

[varargout{1:nargout}] = deep.internal.dlfevalWithNestingCheck(fun,varargin{:}); how to resolve it

Answer 1

[Paras Gupta]

Hi Gaurav,

The error provided in the question indicates that the in-built 'crossentropy' function does not support the argument 'ClassificationMode'. The 'crossentropy' function is being used inside the defined 'classConfidenceLoss' and 'objectnessLoss' functions of the example 'Lidar Object Detection Using Complex-YOLO v4 Network'.

It is to be noted that the argument 'ClassificationMode' was introduced in MATLAB R2023b. The R2023a equivalent of this argument is 'TargetCategories'. You can refer to the 'Version History' section of the following documentation for more information on the same:

https://www.mathworks.com/help/deeplearning/ref/dlarray.crossentropy.html

Please refer to the following version-specific documentation of the example for MATLAB R2023a which uses the correct argument for 'crossentropy':

https://www.mathworks.com/help/releases/R2023a/lidar/ug/lidar-object-detection-using-complex-yolov4.html

Hope this helps resolve your query.