Bubble velocimetry using the conventional and cnn-based optical flow algorithms

HIGHLIGHTS

• who: Daehyun Choi from the DepartmentNational University have published the research work: Bubble velocimetry using the conventional and CNN-based optical flow algorithms, in the Journal: Scientific Reports Scientific Reports of 30/03/2022
• what: As a promising solution to this problem, the authors propose the optical flow algorithm as a bubble velocimetry method. The model is trained following the learning schedule, as introduced in Sun et_al35, which reduces the learning rate (initially starting from u00ad10-5) by half at 160 k, 240 k, 320 k, and 400 k epochs, with the batch size of . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or E-mail:

  If you don't have an account, you can create one here.