DCTNet
DCTNet copied to clipboard
Clarification about dataset
In the instructions, you mention that the ImageNet dataset should be present in the 'data' directory? Does this mean, that I have to download the entire Imagenet dataset to be able to run your script?
For example, what can I do if I want to run inference on a single image with your DCT-24 model, akin to running a trained Resnet-50 for inference on an image?
Can you please respond to this? Can I proceed trying out your code without downloading the ImageNet repository?
You need to download the entire ImageNet dataset if you want to use our code without any modifications. But if you want to test on a single image, you may need to write a customized dataloader. Thanks.
Thank you for the response. I understand. Can you point me which method(s) I may need to modify to run the inference on a single image?
Hi @calmevtime, Can you please let us know what changes to be made exactly in order to perform inference using your pre-trained model on a custom image.
Q1. Do we perform the whole pre-processing on the test image as well just like you do in the validation dataset or just directly input the image to the Model ?
Hi @suvatojo , were you able to get any solution that you can share please?
@SarfarazHabib - No, I have been unable to make any progress on this. I need to try this approach for inference on a single image but I haven't been able to look into the data loaders as @calmevtime suggested. Please share if a way can be found to run inference on a single image.