First of all, Thanks for your excellent paper. I have a question regarding the baseline, particularly FRoST, which relies on prior knowledge such as the number of novel classes. The paper highlights that CGCD cannot accommodate any prior knowledge about novel classes. How did you address this to ensure a fair comparison when implementing FRoST?
Hello Author
First of all, Thanks for your excellent paper. I have a question regarding the baseline, particularly FRoST, which relies on prior knowledge such as the number of novel classes. The paper highlights that CGCD cannot accommodate any prior knowledge about novel classes. How did you address this to ensure a fair comparison when implementing FRoST?
Thank you in advance.