search

Ha0Tang / SelectionGAN

[CVPR 2019 Oral] Multi-Channel Attention Selection GAN with Cascaded Semantic Guidance for Cross-View Image Translation
http://disi.unitn.it/~hao.tang/project/SelectionGAN.html
460 stars 63 forks source link

inception score #7

Closed 04lm40 closed 4 years ago

04lm40 commented 4 years ago

Hello, I trained a model on the CVUSA dataset, but the IS value and KL result are slightly worse during the evaluation. The following are my running commands. Are there any problems that I have not noticed? python train.py --dataroot ./datasets/cvusa/ \ --name cvusa_selectiongan \ --model selectiongan \ --which_model_netG unet_256 \ --which_direction AtoB \ --dataset_mode aligned \ --norm batch \ --gpu_ids 0,1 \ --batchSize 4 \ --loadSize 286 \ --fineSize 256 \ --no_flip \ --display_id 1 \ --lambda_L1 100 \ --lambda_L1_seg 1

Ha0Tang commented 4 years ago

What are your results? Did you use two GPUs for training?

04lm40 commented 4 years ago

IS：3.5 2.65 3.59 KL：3.12. Yes, I used two GPUs.

Ha0Tang commented 4 years ago

The results in our paper were generated using one GPU.

04lm40 commented 4 years ago

Thank you!