search

javiribera / locating-objects-without-bboxes

PyTorch code for "Locating objects without bounding boxes" - Loss function and trained models
Other
249 stars 52 forks source link

training from scratch for the ShanghaitechB dataset #14

Closed 645873229 closed 4 years ago

645873229 commented 4 years ago

I use the pre-trained model you gave (ShanghaitechB),and get a pretty good results ,but I use the example you gave about train.py, and try to train the dataset (ShanghaitechB). But I get totally different results. Your pre-trained get good results, but my model can not get good results. the esitimated_map are full of yellow . It can not locate anyone! No matter which epoch I choose(100,500,1000),results are same. Full of yellow blocks

javiribera commented 4 years ago

I don't know which example you are talking about. My example in the readme file is not meant to be used to train any neural network or any dataset in particular, but only to instruct on which parameters one can provide to the script. Training a neural network is not trivial task and it takes a while to experiment with hyperparameters.

javiribera commented 4 years ago

Closing due to inactivity. Please describe your question more clearly.