The training setting for the CycelGAN

naoto0804 / cross-domain-detection

Cross-Domain Weakly-Supervised Object Detection through Progressive Domain Adaptation [Inoue+, CVPR2018].

https://naoto0804.github.io/cross_domain_detection/

421 stars 77 forks source link

The training setting for the CycelGAN #34

Closed Leesoon1984 closed 4 years ago

Leesoon1984 commented 4 years ago

Can you provide more details about the training process of CyceGeAN? During the image translating form VOC to Clipart, did you use all the clipart1k images or just the train set(not including the test dataset)?

naoto0804 commented 4 years ago

Train set only.

Leesoon1984 commented 4 years ago

Thank you for your prompt reply.