Hello, I found a performance issue in the definition of testResizeToRangeWithDynamicSpatialShape, core/preprocessor_test.py, tf.shape(out_image) will be created repeatedly during program execution, resulting in reduced efficiency. I think out_image_shape and in_image should be created before the loop.
Looking forward to your reply. Btw, I am very glad to create a PR to fix it if you are too busy.