Is this okay that you validate on test set?

gjy3035 / GCC-SFCN

This is the official code of spatial FCN in the paper Learning from Synthetic Data for Crowd Counting in the Wild [CVPR2019].

https://gjy3035.github.io/GCC-CL/

MIT License

160 stars 39 forks source link

Is this okay that you validate on test set? #25

Open BorisLestsov opened 5 years ago

BorisLestsov commented 5 years ago

https://github.com/gjy3035/GCC-SFCN/blob/a5921d2b1e4a6aaa3019568bc4e4fca9d386970e/loading_data.py#L33 - here validation set is created from test set of UCF_QNRF. Is this okay?

gjy3035 commented 5 years ago

we will add the following description: In this code, the validation is directly on the test set. Strictly speaking, it is evaluated on the val set (randomly selected from the training set, which is adopted in the paper). Here, for a comparable reproduction (namely fixed splitting sets), this code directly adopts the test set for validation, which causes that the results of this code are better than that of our paper. Thanks a lot!