Junyu Gao

The link for translation images is: http://share.crowdbenchmark.com:2443/home/Translation_Results

Shanghai B is a small-scale dataset, the model is prone to overfitting. I suggest you train a new model on your data or train a pre-trained model on a large-scale...

Use https://github.com/gjy3035/C-3-Framework/blob/python3.x/datasets/QNRF/preapre_QNRF.m 投process SHHA.

@muzishen https://mailnwpueducn-my.sharepoint.com/:f:/g/personal/gjy3035_mail_nwpu_edu_cn/EkxvOVJBVuxPsu75YfYhv9UBKRFNP7WgLdxXFMSeHGhXjQ?e=IdyAzA

@ydlu 上海B自身尺寸就是768x1024，因此不会有任何变化. 量化后准确性基本不会变化。 @muzishen We will reproduce the generation of the density map from the dot map or points using conv layer. So we have no plan to provide the...

@ydlu 抱歉，看错了你的评论。使用prepare_QNRF.py去resize即可。参见：https://github.com/gjy3035/C-3-Framework/issues/54#issuecomment-563015061

Thanks for your attention. My original goal is for effective reproduction. I will add some explanations for this problem.

pred instead of sum(pred_map)

Thanks a lot! It is a bug and `mse` should be replaced with `0`. We will fix it ASAP.

I do not find this problem. Please check your part4. Or, download it from OneDrive. Here, I upload 3746, and you can add it to your test set. Thanks for...