search

leeyeehoo / CSRNet-pytorch

CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes
642 stars 259 forks source link

Training-Validation split #11

Open BedirYilmaz opened 5 years ago

BedirYilmaz commented 5 years ago

Hello

Thank you for sharing such beautiful work.

May I ask which experimental setup you have used to achieve the results stated in the paper?

I saw that you have two different setups in your json files. I would be glad if I could find out which one is the official.

Did you use validation? If yes, how many of the training images have you reserved for validation set?

leeyeehoo commented 5 years ago

I'll update the repo soon... Sorry for the delay. Besides what you mean the two different setups?

BedirYilmaz commented 5 years ago

There is

part_A_train_with_val.json and part_A_train.json.

In latter, you have he paths of 300 images. So if you use that during training, you will apply no validation I guess.

But in first file you have 240 Images which makes me think that you have reserved 60 images for validation during training. In first glance, these two looked like they belong to two different setups to me. Am I wrong?