search

jhoffman / cycada_release

Code to accompany ICML 2018 paper
BSD 2-Clause "Simplified" License
561 stars 126 forks source link

Reproducing paper results #3

Closed engharat closed 6 years ago

engharat commented 6 years ago

Hi, I would like to replicate the paper results ( SVHN->MNIST for example) but I just noticed that the test_cycada.sh script and the following test.py produce just the converted images and do not produce any accuracy at all. Could you suggest me how to operate in order to reproduce the numerical paper results? Should I write my own test function, testing the target dataset after a full training?

jhoffman commented 6 years ago

The accuracy is reported from the final model which performs feature adaptation. See scripts/train_adda.py for an example which trains directly with svhn -> mnist (feature adapt only) to include image adaptation as well use the dataset src='svhn2mnist' and point to the folder which is the output from image adaptation or directly download the relevant images here: https://eecs.berkeley.edu/~jhoffman/cycada/svhn2mnist/