search

sc0ty / subsync

Subtitle Speech Synchronizer
GNU General Public License v3.0
1.31k stars 57 forks source link

Dictionaries and speech recognition models requests #85

Open sc0ty opened 4 years ago

sc0ty commented 4 years ago

This is aggregated issue to request support for new languages. If you see one of the following errors:

Synchronization between languages xxx - yyy is currently not supported.

Synchronization with xxx audio is currently not supported.

Instead of opening new ticket, just write comment here and I will add it to the list.

Dictionaries:

Speech recognition models:

If you want to help by creating assets (requested here or not) see here for some technical description. All help is appreciated.

fawzib commented 4 years ago

couldn't you change the frequency range and just keep vocals and check when subtitles start and end and try to shift the subtitles to match by display length of the subtitle and audio length. this might be a very generic approach

sc0ty commented 4 years ago

There is another project that does exactly that. I'm planning to do something similar eventually, but my synchronizer architecture would poorly fit this approach, so I will need to implement separate synchronization engine. That means lots of work, so don't expect it to be done in the near future.

hista commented 4 years ago

A comment to ask if you could add French language support :-)

sc0ty commented 4 years ago

Added to the list.

abelrod666 commented 4 years ago

It will be great to have Japanese and korean recognition. There is stll a chance to get them ? Thank you for your app. It's very useful.

sc0ty commented 4 years ago

Sorry @abelrod666, I've missed your reply.

Subsync is using Sphinx speech recognition engine with language models taken from the internet. I don't have knowledge to make new models, and there is no publicly available models for languages on this list (that I know of). That's why we don't support them. If you know of any missing models or you are able to create one then I can add it to subsync, otherwise I can't help you.

abelrod666 commented 4 years ago

@sc0ty No problem. I understand. Gonna see if I can find it.... Thank you for your app. It's great and saves me a lot of time.

srmajid commented 3 years ago

please add persian language support

K0ng2 commented 2 years ago

please add Thai language support too.

kid1485 commented 2 years ago

please add Vietnamese language support too

Dnkhatri commented 2 years ago

The new openai whisper look really interesting specailly as it can detect multiple languages and isolate words from noise.

https://github.com/openai/whisper

fuatsarperasli commented 1 year ago

please add Turkish language support (your subsync is exellent!)

ronohkmo commented 11 months ago

please add Cantonese support

doththouevenhoist commented 10 months ago

Any possibility of adding Ukrainian support? Thank you for all your work on this!

JanikSi commented 8 months ago

Please add Polish. Thanks in advance.

JanikSi commented 8 months ago

Hello, is it possible to add speech-cze and speech-pol. Thanks in advance.

KayJannOnGit commented 4 months ago

Hi, your job is Amazing I love it but please I need Japanese Speech recognition model ! thanks in advance

veselinve commented 2 months ago

Is Japanese Speech recognition model available somewhere so we can add it ?

veselinve commented 2 months ago

I've found only this. But i don't know, if it's the right one: https://huggingface.co/jonatasgrosman/wav2vec2-large-xlsr-53-japanese