I've trained a FasterRCNNObjectDetector as shown in the code below:
Tabulate your labeled growth truth data into a table
data4=load('labelsArrayLettucePrev.mat');
table1=table(data4.gTruth.DataSource.Source,data4.gTruth.LabelData.lettuce);
ObjectLoc=table1;
Training Option
% Training Option
options = trainingOptions('sgdm', ...
'MiniBatchSize', 1, ....
'InitialLearnRate',1e-4, ...
'MaxEpochs',13,...
'CheckpointPath', tempdir, ...
'executionEnvironment','gpu',...
'Shuffle','every-epoch');
Train Object Detector
% Train Faster RCNN detector.
[detector, info] = trainFasterRCNNObjectDetector(ObjectLoc, "alexnet", options);
This gives me a .mat that i want to export as a tflite so that i can use my object detector in flutter with flutter_tflite, since the code is already built for a tflite model.