Conformer
Conformer copied to clipboard
object detection code
I saw the object detection result in the paper, but the input image size is (800,1333), I think may cause the model take up more space. Is there any code to achieve object detection?