The train pipeline for CARPK dataset

Question

The train pipeline for CARPK dataset

shanguangcv opened this issue 2 years ago · 3 comments

Hi, in cross-dataset (CARPK) experiment, you pre-train the model on FSC-147, can you provide a pre-train script? Moreover, can you also provide a fine-tune script for easy use? Thanks.

shanguangcv commented 2 years ago

Thanks

Answer 1 · 2022-09-26T03:35:04.000Z

Hi~

This needs some time to re-organize.

The main reason is that the settings of FSC147 & CARPK are different: for FSC147, exemplars are parts of the query image, for CARPK, exemplars are sampled from the training set and fixed for both training and evaluation.

We hope provide this in this week.

Answer 2 · 2022-10-02T16:28:46.000Z

Hi~
We have provided the pre-train and fine-tune script. If you have other questions, please re-open this issue.