search

mecarill / 2pcnet

Other
85 stars 22 forks source link

backbone model #3

Closed achintski closed 1 year ago

achintski commented 1 year ago

Dear sir, Thank you for sharing your work！ I have a question about how to modify the backbone model.The backbone used in your code is Faster RCNN, and I want to modify it to Yolov7. Since I am just getting started with object detection, I am not familiar with modifying the code. I kindly ask you to provide some ideas for modification (backbone part and train part), and I will modify it by myself according to your ideas. Looking forward to hearing from you！ Thanks in advance！

mecarill commented 1 year ago

Hello achintski! Due to the architecture of this current method, it would not be simple to copy/paste it into YOLO. This is because we make use of the RPN which is the first part of the 2-stage object detector (FRCNN) whereas YOLO is 1-stage.

You could potentially use the outputs of the teacher in YOLO as a proxy for the RPN outputs but we cannot guarantee performance as we have not tested it.