Tutorial error: Unable to apply MiniBatchFormat value 'SSCB' to output 1 on Yolo V3 Deep Learning

I am following the MATLAB Tutorial: Object Detection Using Yolo V3 Deep Learning: https://uk.mathworks.com/help/vision/ug/object-detection-using-yolo-v3-deep-learning.html

Whenever I get to the model training step, I try to instantiate a minibatchqueue object. This is the code:

mbqTrain = minibatchqueue(preprocessedTrainingData, 2,...
        "MiniBatchSize", miniBatchSize,...
        "MiniBatchFcn", @(images, boxes, labels) createBatchData(images, boxes, labels, classNames), ...
        "MiniBatchFormat", ["SSCB", ""],...
        "DispatchInBackground", dispatchInBackground,...
        "OutputCast", ["", "double"]);

Although I have been following this tutorial exactly (haven't changed any parameters, am using the example dataset, etc.) I keep getting the following error:

Error using minibatchqueue (line 319)
Unable to apply MiniBatchFormat value 'SSCB' to output 1.

Caused by:
    Error using dlarray (line 145)
    Invalid data type. Arguments must be full double, single, or logical arrays.

I am also using the pretrained detector they use and have re-downloaded it several times.

I'm currently running MATLAB 2020b (updated).

[Kenza Benlamlih]

did u solve ur problem

[Brian Watson]

Just tried it in 2021a and it still fails. Sorry I don't have an answer.

Answer 1

[Brian Watson]

This YOLO example must be run in Matlab version 2021a or it fails.