Hi,
Very interesting work, we are eager to reproduce it!
But we found no build_batches.py. We tried to use a file with the same name in TF-phrasecut-public, but didn't get the same dataset structure as Readme under this repository (We only get the batch folder, not image and label.).
We are very worried that the data processing is different from yours. If it is convenient could you provide the missing build_batches.py.
Best Regards.
Hi,
Very interesting work, we are eager to reproduce it!
But we found no build_batches.py. We tried to use a file with the same name in TF-phrasecut-public, but didn't get the same dataset structure as Readme under this repository (We only get the batch folder, not image and label.).
We are very worried that the data processing is different from yours. If it is convenient could you provide the missing build_batches.py.
Best Regards.