Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

OTranscribe: A free and open tool for transcribing audio interviews

otranscribe.com

Post reply on HN
Print view View original

101–109 of 109 posts

Re: OTranscribe: A free and open tool for transcribing audio interviews

#101

Post by moritzwarhier »

tkgally wrote: post #45

I was curious how good a transcription I could get from what may be the best multimoldal LLM currently, Gemini-1.5-Pro-Experiment-0801, so I had it transcribe five minutes of an interview between Ezra Klein and Nancy Pelosi from earlier today. The results are here: https://www.gally.net/temp/20240809geminitranscription/index... Aside from some minor punctuation and capitalization issues, Gemini’s transcription looks…

Maybe two factors helped achieve the impressive result with the quotation marks:

- auditory cues

- the sentence would be gramatically incorrect and make no sense without them

Just guessing out of the blue.

But I think it's likely that LLMs (and other speech recognition systems) need to exploit sentence context to recognize individual words and punctuation, and this is an example were it went well.

Human listening is similar in a way, we can recognize words even when spoken very mumbly or fast, if we have context.

So we always hear phrased rather than words.

Re: OTranscribe: A free and open tool for transcribing audio interviews

#102

Post by rpdillon »

andrei-akopian wrote: post #67
btown wrote: post #57

Are there any open-source or paid apps/shareware/freeware that can: - Transcribe word-by-word in real time as audio is recorded - Work entirely locally - Use relatively recent open-source local models? I've been using otter.ai for real-time meeting transcriptions - letting me multitask and instantly catch up if I'm asked a question by skimming the most recent few seconds worth of the transcript - but it's far from pe…

futo.org has FOSS voice input android app (voiceinput.futo.org) and live captions ( https://github.com/abb128/LiveCaptions ) for Linux. They specifically developed their own model that does fast real time transcriptions. Not sure if that helps for your specific usecase.

I've been using this for the past two or three weeks and I have been very impressed with it. I ended up giving them five dollars because it's now my primary keyboard!

Re: OTranscribe: A free and open tool for transcribing audio interviews

#103

Post by windthrown »

leiferik wrote: post #54

You're always welcome to try my service TurboScribe https://turboscribe.ai/ if you need a transcript of an audio/video file. It's 100% free up to 3 files per day (30 minutes per file) and the paid plan is unlimited and transcribes files up to 10 hours long each. It also supports speaker recognition, common export formats (TXT, DOCX, PDF, SRT, CSV), as well as some AI tools for working with your transcript.

Thanks! I've had good results with Turboscribe (paid plan) and appreciate having this as a service. I typically use it for 2-3 hour long video recordings with a number of speakers and appreciate the editing tools to clean things up before export.

Re: OTranscribe: A free and open tool for transcribing audio interviews

#105

Post by sebastiennight »

tkgally wrote: post #45

I was curious how good a transcription I could get from what may be the best multimoldal LLM currently, Gemini-1.5-Pro-Experiment-0801, so I had it transcribe five minutes of an interview between Ezra Klein and Nancy Pelosi from earlier today. The results are here: https://www.gally.net/temp/20240809geminitranscription/index... Aside from some minor punctuation and capitalization issues, Gemini’s transcription looks…

It's very likely that the model is capable of picking up on the verbal cues surrounding quotes.

Do you have the audio or video file?

I'd like to run it through our AI video editor and see how it punctuates the transcript.

Re: OTranscribe: A free and open tool for transcribing audio interviews

#106

Post by jduckles »

terribleperson wrote: post #92
cube2222 wrote: post #5

I needed to do this this week (transcribe an interview with multiple speakers) and used https://github.com/MahmoudAshraf97/whisper-diarization Worked excellent. It generates both a file that just contains a line per uninterrupted speaker speech prefixed with the speaker number, as well as a file with timestamps which I believe would be used as subtitles.

Does it hallucinate when there's dead air?

Yes

Re: OTranscribe: A free and open tool for transcribing audio interviews

#107

Post by tkgally »

sebastiennight wrote: post #105
tkgally wrote: post #45

I was curious how good a transcription I could get from what may be the best multimoldal LLM currently, Gemini-1.5-Pro-Experiment-0801, so I had it transcribe five minutes of an interview between Ezra Klein and Nancy Pelosi from earlier today. The results are here: https://www.gally.net/temp/20240809geminitranscription/index... Aside from some minor punctuation and capitalization issues, Gemini’s transcription looks…

It's very likely that the model is capable of picking up on the verbal cues surrounding quotes. Do you have the audio or video file? I'd like to run it through our AI video editor and see how it punctuates the transcript.

The mp3 file that I gave to Gemini (a five-minute excerpt from the audio podcast) is linked in the source code of the page. Here is the full URL:

https://www.gally.net/temp/20240809geminitranscription/inter...

The full interview including video is on the New York Times website, though you might need a subscription to view it:

https://www.nytimes.com/2024/08/09/opinion/ezra-klein-podcas...

The NYT’s closed captions do not put “weird” in quotation marks; they also divide sentences weirdly and have other mistakes. But they get some things better than Gemini, such as capitalizing “House” when it means the House of Representatives.

I haven’t compared the audio-only podcast version and the video version carefully; it’s possible that parts of the audio were edited or re-recorded for one or the other.

Let us know how your AI video editor does!

Re: OTranscribe: A free and open tool for transcribing audio interviews

#108

Post by sebastiennight »

tkgally wrote: post #107
sebastiennight wrote: post #105

Earlier quoted context omitted.

It's very likely that the model is capable of picking up on the verbal cues surrounding quotes. Do you have the audio or video file? I'd like to run it through our AI video editor and see how it punctuates the transcript.

The mp3 file that I gave to Gemini (a five-minute excerpt from the audio podcast) is linked in the source code of the page. Here is the full URL: https://www.gally.net/temp/20240809geminitranscription/inter... The full interview including video is on the New York Times website, though you might need a subscription to view it: https://www.nytimes.com/2024/08/09/opinion/ezra-klein-podcas... The NYT’s closed captions do…

OK I ran it.

We have a grammar correction step which "fixed" the sentence to "...he comes up with something weird, which becomes viral...".

Re: OTranscribe: A free and open tool for transcribing audio interviews

#109

Post by baby_souffle »

ericjmorey wrote: post #71
baby_souffle wrote: post #61

Earlier quoted context omitted.

> Are there any open-source or paid apps/shareware/freeware Google Pixel phones have this feature and it works _very_ well.

How is that feature accessed? Or what does Google call it so I can search for it.

It's the "recorder" app. It was released in 2021 I think and has been improving since the beginning.

If you elect to, the audio + transcription can be access/searched via https://recorder.google.com

Post reply on HN