天涯古巷
Results
89
comments of
天涯古巷
it maight be you choose the wrong recognition model, you can set recognition language in ```settings.ini``` > e.g. change ```ch``` to ```en```
已经有这个功能了 你subtitle_area传入一个None就可以了
这时候就该用精准模式了
开始的时候提取字幕帧使用的cpu 后面就用gpu了
任务管理器看的不准,要用cmd的nvidia-smi命令才准。新版本提取帧和识别一起做了,提高了运算效率,新版本自动检测GPU,有就开启,默认使用高精度模型
跑的gui还是cli
sudo执行试试
do you mean Turkish interface or Turkish subtitle extraction
Turkish corpus is needed in order to support turkish subtitle extraction. Collecting corpus is a very time-consuming work. I'm sorry I might have no time to do this. This repo...