When will have diarization?

[AuroraMartell]

Could you explain what diarization means?

[johnfelipe]

Put name of speaker

Speaker 1: xxxxxx Speaker 2: yyyyyy Speaker 2: zzz zzz . . . Etc

El mié, 27 dic 2023, 11:16 a. m., Aurora Martell @.***> escribió:

Could you explain what diarization means?

— Reply to this email directly, view it on GitHub https://github.com/SubtitleEdit/subtitleedit/issues/7794#issuecomment-1870445648, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADIWFCY4IFA6XR7TX62O4TYLRCVHAVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ2DKNRUHA . You are receiving this because you authored the thread.Message ID: @.***>

[AuroraMartell]

Put name of speaker Speaker 1: xxxxxx Speaker 2: yyyyyy Speaker 2: zzz zzz .

I think you can use the "Set Actor" feature for this, but I'm not sure since I've never needed to use it.

[johnfelipe]

Can u send me screenshots or video for how use that feature?

El mié, 27 dic 2023, 11:38 a. m., Aurora Martell @.***> escribió:

Put name of speaker Speaker 1: xxxxxx Speaker 2: yyyyyy Speaker 2: zzz zzz .

I think you can use the "Set Actor" feature for this, but I'm not sure since I've never needed to use it.

— Reply to this email directly, view it on GitHub https://github.com/SubtitleEdit/subtitleedit/issues/7794#issuecomment-1870462194, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADIWFCRGLQKSS6FNQCILZDYLRFJ5AVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ3DEMJZGQ . You are receiving this because you authored the thread.Message ID: @.***>

[LeonCheung]

As @AuroraMartell said, you can do this with a trick by using the Actor column of ASSA format:

Say, you want to export it as SRT, then you can go to File > Export > Export custom text format... and create your own SRT format like this:

Also, you can set shortcuts to assign actor/speaker labels faster:

[AuroraMartell]

As @AuroraMartell said, you can do this with a trick by using the Actor column of ASSA format:

Say, you want to export it as SRT, then you can go to File > Export > Export custom text format... and create your own SRT format like this:

Also, you can set shortcuts to assign actor/speaker labels faster:

This is exactly what I was describing! Thanks, @LeonCheung (:

[johnfelipe]

Ahh, but this is manual, diarization is automatically May be can you add this feature in your roadmap?

El mié, 27 dic 2023, 11:41 a. m., Leon Cheung @.***> escribió:

As @AuroraMartell https://github.com/AuroraMartell AuroraMartell said, you can do this with a trick by using the Actor column of ASSA format: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/e6abcba9-539a-4814-b23e-0a0886bbcd1e

Say, you want to export it as SRT, then you can go to File > Export > Export custom text format... and create your own SRT format like this:

image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/7b2bdee8-2470-4101-907d-953777fb92a3

Also, you can set shortcuts to assign actor/speaker labels faster:

image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/a36278d1-1935-406c-bb3c-2ff25be15eba

— Reply to this email directly, view it on GitHub https://github.com/SubtitleEdit/subtitleedit/issues/7794#issuecomment-1870463999, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADIWFAPWRAULBXV6BZXPTTYLRFTJAVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ3DGOJZHE . You are receiving this because you authored the thread.Message ID: @.***>

[LeonCheung]

Ahh, but this is manual, diarization is automatically May be can you add this feature in your roadmap? El mié, 27 dic 2023, 11:41 a. m., Leon Cheung @.> escribió: … As @AuroraMartell https://github.com/AuroraMartell AuroraMartell said, you can do this with a trick by using the Actor column of ASSA format: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/e6abcba9-539a-4814-b23e-0a0886bbcd1e Say, you want to export it as SRT, then you can go to File > Export > Export custom text format... and create your own SRT format like this: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/7b2bdee8-2470-4101-907d-953777fb92a3 Also, you can set shortcuts to assign actor/speaker labels faster: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/a36278d1-1935-406c-bb3c-2ff25be15eba — Reply to this email directly, view it on GitHub <#7794 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADIWFAPWRAULBXV6BZXPTTYLRFTJAVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ3DGOJZHE . You are receiving this because you authored the thread.Message ID: @.>

AFAIK, some AI transcription tools support this already but they can never be accurate enough to me as I've experienced. BTW, do you know any tools that can typically do this well? Just curious.

[johnfelipe]

im use this https://github.com/kaixxx/noScribe, but is very slow and RAM eater

El mié, 27 dic 2023 a las 11:47, Leon Cheung @.***>) escribió:

Ahh, but this is manual, diarization is automatically May be can you add this feature in your roadmap? El mié, 27 dic 2023, 11:41 a. m., Leon Cheung @.

> escribió: … <#m743533593393906008> As @AuroraMartell https://github.com/AuroraMartell https://github.com/AuroraMartell https://github.com/AuroraMartell AuroraMartell said, you can do this with a trick by using the Actor column of ASSA format: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/e6abcba9-539a-4814-b23e-0a0886bbcd1e https://github.com/SubtitleEdit/subtitleedit/assets/494788/e6abcba9-539a-4814-b23e-0a0886bbcd1e Say, you want to export it as SRT, then you can go to File > Export > Export custom text format... and create your own SRT format like this: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/7b2bdee8-2470-4101-907d-953777fb92a3 https://github.com/SubtitleEdit/subtitleedit/assets/494788/7b2bdee8-2470-4101-907d-953777fb92a3 Also, you can set shortcuts to assign actor/speaker labels faster: image.png (view on web) https://github.com/SubtitleEdit/subtitleedit/assets/494788/a36278d1-1935-406c-bb3c-2ff25be15eba https://github.com/SubtitleEdit/subtitleedit/assets/494788/a36278d1-1935-406c-bb3c-2ff25be15eba — Reply to this email directly, view it on GitHub <#7794 (comment) https://github.com/SubtitleEdit/subtitleedit/issues/7794#issuecomment-1870463999>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADIWFAPWRAULBXV6BZXPTTYLRFTJAVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ3DGOJZHE https://github.com/notifications/unsubscribe-auth/AADIWFAPWRAULBXV6BZXPTTYLRFTJAVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ3DGOJZHE . You are receiving this because you authored the thread.Message ID: @.>

AFAIK, some AI transcription tools support this already but they can never be accurate enough to me as I've experienced. BTW, do you know some tools that can typically do this well? Just curious.

— Reply to this email directly, view it on GitHub https://github.com/SubtitleEdit/subtitleedit/issues/7794#issuecomment-1870468325, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADIWFFHZAT7JSEPAZW2ZB3YLRGLHAVCNFSM6AAAAABBEFZ4PCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZQGQ3DQMZSGU . You are receiving this because you authored the thread.Message ID: @.***>

[niksedk]

When/if it's available in Whisper, it will be available in SE too.

[raphaelomoreira]

When/if it's available in Whisper, it will be available in SE too.

@niksedk hi!! It seems to already be in some Whisper projects. I was able to make this caption through the command prompt (whisper-ctranslate2). These functions also seem to already exist in some projects on SubtitleEdit's Whisper, but I couldn't run them because they need the token on huggingface.co. WhisperX also already has Diarization in the project, but it does not have a parameter field in SubtitleEdit. Do we have any progress in Faster-Whisper?

[johnfelipe]

with faster whisper have same parameter?