I am performing transfer learning on the pretrained ‘swin_tiny_patch4_window7_224’ model, but my input image size is 100x100x3, whereas the model input size is 224x224x3.
How can I modify the architecture so I can use it for 100x100x3 input images?
I am performing transfer learning on the pretrained ‘swin_tiny_patch4_window7_224’ model, but my input image size is 100x100x3, whereas the model input size is 224x224x3.
How can I modify the architecture so I can use it for 100x100x3 input images?