faster RCNN layer warning

3 views (last 30 days)
ahmad
ahmad on 3 Oct 2023
Answered: T.Nikhil kumar on 9 Oct 2023
hi everyone.
i have problem while training my FaterRcnn .when i create FasterRCNNlayer and analyse layer is show warning although the FaterRCNN trained .i want to know shloud i ignore this warning .warning image attached.And my code is
net=resnet101();
featureLayer = 'res5c_relu';
lgraph =fasterRCNNLayers(inputSize,numClasses,anchorBoxes,net, featureLayer);
analyzeNetwork(lgraph)

Answers (1)

T.Nikhil kumar
T.Nikhil kumar on 9 Oct 2023
Hey Ahmad,
I understand that you are trying to create a faster R-CNN Object detection network using ‘fasterRCNNLayers’ function and are facing a warning that says that the network architecture is not supported by ‘trainNetwork’ function.
This warning is to inform you that the newly created faster R-CNN network cannot be trained using the ‘trainNetwork’ function since this network’s architecture is not compatible with it. Instead, you can use the ‘trainFasterRCNNObjectDetector’ function to train this network as a ‘layerGraph’. Therefore, unless you use the ‘trainNetwork’ function there should be no future errors related to this warning.
You can refer to the following documentation for guidance on implementing the trainFasterRCNNObjectDetector function.
Hope this resolves your query!

Community Treasure Hunt

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

Start Hunting!