This is a great job! I have some questions regarding the composition of the REBlur dataset. The method of data composition described in the paper for two shots seems to apply only to scenarios where an action camera captures a static scene. Could you please explain how the paired blurred-sharp data is obtained in scenarios where a handheld calibration board is moved?
This is a great job! I have some questions regarding the composition of the REBlur dataset. The method of data composition described in the paper for two shots seems to apply only to scenarios where an action camera captures a static scene. Could you please explain how the paired blurred-sharp data is obtained in scenarios where a handheld calibration board is moved?