Subtitles. Translate To English without Speech Recognition?

Help
1

Scenario:

  • I have the video. I have already imported a subtitle for it. It appears on the timeline. Everything works.

Questions:

  • Is it possible to “Translate to English” this foreign (Indonesian) subtitle without using the Speech Recognition feature? In other words, using the Subtitle file itself as the source and converting it straight to English (much quicker)?
  • Also, is it possible to integrate 2 subtitles (foreign and English) for a video in kdenlive?

Thank you.

2

A definite No to your first question. Speech recognition has it in its name that it only works on spoken words and the AI models are built around that concept.

AFAIK, there is only one subtitle track available at the moment. Perhaps a capability of more than one is in the works but I don’t know.
If you need a second track use a special subtitle tool (google is your friend) or Handbrake (works with SRT files).

1 Like
3

I don’t know if this it outdated, but currently you can go under Project > Subtitles > Manage Subtitles. I believe here is where you can add more than one subtitle track.

Plus, Facebook Meta’s blog on SeamlessM4T (which is the model integrated in kdenlive for speech recognition) mentions that it supports various inputs and ouputs, including Text to Text translation!

Today, we’re introducing [SeamlessM4T], a foundational multilingual and multitask model that seamlessly translates and transcribes across speech and text. SeamlessM4T supports:

  • Automatic speech recognition for nearly 100 languages
  • Speech-to-text translation for nearly 100 input and output languages
  • Speech-to-speech translation, supporting nearly 100 input languages and 35 (+ English) output languages
  • Text-to-text translation for nearly 100 languages
  • Text-to-speech translation, supporting nearly 100 input languages and 35 (+ English) output languages

Still, even if the “model” / “ai”? Is capable of this task I don’t know yet if it is possible to do something other than speech to text , or speech to translation inside of kdenlive itself. If someone knows how please let us now!