I would like to suggest ways to improve the software's Vietnamese language recognition accuracy.

[jjigoku]

I used this software, which seems to be based on VideoSubfind, but I felt that the subtitle scanning speed was much slower, even in the fast mode. I also commented that the Vietnamese recognition results were extremely poor and unusable. I think you should do it like VideoSubfind: create subtitle images and use Google Drive to recognize the Vietnamese, then use this software to combine them. Using Google Drive for recognition still gives much better results.