Open qianyuzqy opened 3 years ago
Dear Chaoqi, @chaoqichen
Thanks for your nice work of HTCN published in CVPR 2020! I found the paper quite interesting and your clear and compact illustrations inspired me a lot. I use the code you provided to train a new model from Cityscapes to Foggy cityscapes, but got a poor performance. The main reason I guess is that I did not use the translated images of Cityscapes. So could you please send me a copy of your pre-trained model of CycleGAN in City->Foggy or send me your translated dataset? I can use the Baidu WANGPAN and I find your link provided in GitHub is outdated. I will appreciate it if you can take time out of your busy schedule to read my message.
Looking forward to your reply.
Best wishes to you!
Hi Qianyu666, May I have your connection to ask some questions about this paper? Hope for your reply. And I am in SJTU, too.
ZSY
Dear Chaoqi, @chaoqichen Thanks for your nice work of HTCN published in CVPR 2020! I found the paper quite interesting and your clear and compact illustrations inspired me a lot. I use the code you provided to train a new model from Cityscapes to Foggy cityscapes, but got a poor performance. The main reason I guess is that I did not use the translated images of Cityscapes. So could you please send me a copy of your pre-trained model of CycleGAN in City->Foggy or send me your translated dataset? I can use the Baidu WANGPAN and I find your link provided in GitHub is outdated. I will appreciate it if you can take time out of your busy schedule to read my message. Looking forward to your reply. Best wishes to you!
Hi Qianyu666, May I have your connection to ask some questions about this paper? Hope for your reply. And I am in SJTU, too.
ZSY
Hi, ZSY, Sorry for the late reply. I did not notice the message here. Do you still have any questions? You can contact me at zhouqianyu@sjtu.edu.cn.
Dear Chaoqi, @chaoqichen
Thanks for your nice work of HTCN published in CVPR 2020! I found the paper quite interesting and your clear and compact illustrations inspired me a lot. I use the code you provided to train a new model from Cityscapes to Foggy cityscapes, but got a poor performance. The main reason I guess is that I did not use the translated images of Cityscapes. So could you please send me a copy of your pre-trained model of CycleGAN in City->Foggy or send me your translated dataset? I can use the Baidu WANGPAN and I find your link provided in GitHub is outdated. I will appreciate it if you can take time out of your busy schedule to read my message.
Looking forward to your reply.
Best wishes to you!