nice to meet you

[12412414134113422]

Hello, nice to meet you. I am a first-year graduate student from BIT. May I ask if there is a corresponding paper for this article?
I want to learn from you to discuss.Thank you

[CesareMagnetti]

Hi, thanks for reaching out! Unfortunately there is no paper about this specific repo, this work was done as part of the AutomaticUSnavigation project (https://github.com/CesareMagnetti/AutomaticUSnavigation) but the models were not working great when sampling CT planes on the fly. The main blocker was the lack of a good quality US dataset containing non standard views. As a consequence our models would always produce US images that looked like standard planes even when they shouldn’t. Most of the work in this repo simply uses CycleGAN and CUT from their original repositories, the only contribution was using LPIPS as a cycle consistency loss. This work was part of my MSc thesis, I can send you that if you need it!

[12412414134113422]

Thank you so much! If you don't mind, you can send me a copy, and I can also learn, you can send to my email: carbeen_cao@163.com. I see the result of your project is still very good, I am also learning and doing simulation generation, but the effect is not satisfactory. If possible, I can communicate with you more in the future. I am in Beijing Institute of Technology, China. If you have any needs, please feel free to contact me.
And then I have one more question for you. May I ask, did you directly train and test on Cyclegan and CUT? How did you solve the data problem?
Also, do the data input into the training need to be paired with CT and US?
thank you

[12412414134113422]

Hello, dear senior. Excuse me, could you please send me your pre-training model so I want to study it? My mainframe is too weak and training takes too long. Besides, has your master's thesis been published? May I study it? Thank you so much~