Closed cy-yong closed 2 years ago
Hi @cy-yong,
This problem is likely related to the image size. For a 3-level LapIRN model, the input size in each dimension should be divisible by 2^4=16, e.g.: (144, 160, 192).
You may consider crop/pad-zero/resize for your image. See the example in Train_LapIRN_diff_resize.py if you want to resize your image.
当我用我自己的数据集的时候我遇到了这样的问题,请问您能告诉我这个问题可以怎么解决吗?