After scaling the input image to a different resolution, is it possible to keep the original graph (used to detect bounding boxes on the original sized image).
Or do i have to load a new graph to detect the bounding boxes on the image with different resolution?
After scaling the input image to a different resolution, is it possible to keep the original graph (used to detect bounding boxes on the original sized image). Or do i have to load a new graph to detect the bounding boxes on the image with different resolution?
Thanks!