search

wuhuikai / FastFCN

FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation.
http://wuhuikai.me/FastFCNProject
Other
838 stars 148 forks source link

Training on custom dataset #19

Closed loveis98 closed 5 years ago

loveis98 commented 5 years ago

Hello! Will 12 GB of single GPU be enough to train your model on a small dataset? And may be you can publish your pretrained model (dataset no matter)? It would be great)

wuhuikai commented 5 years ago
  1. The pre-trained model can be found here.
  2. A single GPU is enough if you freeze all the BN layers.
meanmee commented 4 years ago

@wuhuikai could you show us how to make a custom dataset to fit into the training process? what is the data structure?

wuhuikai commented 4 years ago

Please refer to the Base class and a concrete class.