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

Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

news.ycombinator.com

Post reply on HN
Print view View original

221–229 of 229 posts

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#221

Post by ChrisGammell »

audiohermit wrote: post #38

Hey, speech ML researcher here. Make sure you have different recordings of different contexts. fifteen.ai's best TTS voices use ~90 min of utterances, some separated by emotion. If you're having her read a text, make sure it's engaging--we do a lot of unconscious voicing when reading aloud. Tbh, if she has a non-Anglophone accent, you're going to need more because the training data is biased towards UK/US speakers. I…

I have about 500 hours of high quality, channel isolated (separate from the person I was speaking to) audio. It comes from my podcast that I have done for many years. It's probably closer to 75-100 hours audio of me actually speaking, since I am more the interviewer.

Is that something that would be useful to a researcher in any context? I am intrigued by the idea of having my voice preserved (you know, ego), but also am happy to donate the sound files if they would help researchers in any way for datasets.

If so: chris@theamphour.com

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#222

Post by erogol »

Hope I am not repeating any comments here. My suggestion is that you start recording as soon as possible and as much as possible without worrying about technicalities. You can also use if you have any old voice records or videos with a relatively good voice quality. For now maybe she can read a book aloud in a silent room. After you have the data I can also help if you like to create a TTS model.

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#223

Post by lunixbochs »

exikyut wrote: post #213
lunixbochs wrote: post #167

Earlier quoted context omitted.

I have an open source web service for rapidly recording lots of text prompts to flac: https://speech.talonvoice.com (right now the live site prompts for single words because I’m trying to build single word training data, but the prompts can be any length) You can set it up yourself with a bit of Python knowledge from this branch: https://github.com/talonvoice/noise/tree/speech-dataset There are keyboard shortcuts - u…

Somewhat OT question: after taking a quick look at this I stumbled on the eye-tracking video you made using pops to click... and I'm curious, can eye trackers not detect and report blinking? Also, I noted the VLC demo says it doesn't use DNS! That's awesome...

I can detect blinking, yes. However your eyelids have very small muscles that are not meant to be consciously controlled all day and I don’t recommend straining them. Your eyelids twitching from muscle strain is rather uncomfortable (from experience)

The VLC demo was using macos speech recognition. In the beta now I’m shipping my own engine+trained models based on Facebook’s wav2letter, which is going pretty well.

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#224

Post by lunixbochs »

ChrisGammell wrote: post #221
audiohermit wrote: post #38

Hey, speech ML researcher here. Make sure you have different recordings of different contexts. fifteen.ai's best TTS voices use ~90 min of utterances, some separated by emotion. If you're having her read a text, make sure it's engaging--we do a lot of unconscious voicing when reading aloud. Tbh, if she has a non-Anglophone accent, you're going to need more because the training data is biased towards UK/US speakers. I…

I have about 500 hours of high quality, channel isolated (separate from the person I was speaking to) audio. It comes from my podcast that I have done for many years. It's probably closer to 75-100 hours audio of me actually speaking, since I am more the interviewer. Is that something that would be useful to a researcher in any context? I am intrigued by the idea of having my voice preserved (you know, ego), but also…

Do you have transcripts, even just for some of the episodes? Unsupervised learning is possible but more difficult.

In general, yes, this is probably useful data in some way for speech recognition or TTS.

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#225

Post by lunixbochs »

daanzu wrote: post #217
audiohermit wrote: post #38

Hey, speech ML researcher here. Make sure you have different recordings of different contexts. fifteen.ai's best TTS voices use ~90 min of utterances, some separated by emotion. If you're having her read a text, make sure it's engaging--we do a lot of unconscious voicing when reading aloud. Tbh, if she has a non-Anglophone accent, you're going to need more because the training data is biased towards UK/US speakers. I…

I wrote a simple little Python GUI app to record training audio. Given a text file containing prompts, it will choose a random selection and ordering of them, display them to be dictated by the user, and record the dictation audio and metadata to a .wav file and recorder.tsv file respectively. You can select a previous recording to play it back, delete it, and/or re-record it. It comes with a few selections of senten…

Did you figure out why half of Shervin’s audio was empty? I would hesitate to recommend this if there’s still a chance half of the data isn’t usable after recording.

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#226

Post by unstatusthequo »

hvaoc wrote: post #208

This is not open source but this was very good from their demo in terms of your own voice reproduction. https://www.descript.com/lyrebird-ai I hope good folks in there will help you, try reaching them. https://m.youtube.com/watch?v=VnFC-s2nOtI

Love Descript and think it’s a great way to both record and get transcripts.

Re: Ask HN: My wife might lose the ability to speak in 3 weeks – how to prepare?

#227

Post by maps7 »

ardenwood wrote: post #190

Hi, I like your idea for your wife. Hope the surgery will succeed without damage to her speaking. I'm from Nvidia and know well the team behind NeMo toolkit. Happy to connect you to the team if that helps. You may send me an email to ardenwood.bruin_at_gmail.com. -- Michael

That's really good of you. It's amazing to see this community be so helpful.
Post reply on HN