Mount already trained ResNet classifcation model at the end of detectron bbox output?

[lehigh123]

I have a ResNet50 image classification model I've trained (in Pytroch) to identify different types of airplanes.

Is there some way I can use a pre-trained detectron2 model (trained on COCO) and use it to identify the bounding box of airplanes (which is already a COCO class) and feed the resulting image within the airplane's bounding box into my already trained ResNet model?

Currently, I'm running my images through detectron2, getting the bounding boxes for airplanes, and then feeding those images into my ResNet model. I am hoping there is a way to avoid this step.