had an interesting project request come across my desk.
A #deafblind client can no longer listen and is missing out on talk radio. how easy would it be to set up a receiver that output text they could read in Braille?
I can presumably get a raspberry pi to start streaming inbound audio either from an SDR or the web. is it viable to use something like #Whisper to then pipe that out somewhere?
I don't want to necessarily store data, although that could be useful I suppose, but the real aim is to have the transcript as close to real time as possible coming out over a local network connection (telnet, probably, for prototyping).
Anyone have any experience of this?

People are seriously criticizing the integration of whisper into VLC? It's an accessibility feature that will benefit everyone, and with little impact if not perfect. It runs on device and is exactly the type of improvement we'd expect from reasonable AI systems.
The only issue I see is that the training data isn't open source, but the model and inference code is.
#VLC #whisper #OpenAI

Johnny5 - #XMPP #Whisper #bot
cyberdelia.com.ar/johnny5-xmpp…

@Mara It's deepl.com. At least I think this is what @Paweł Masarczyk mentioned.
More about #whisper. I like the fact it can all be run locally.

How accurate is transcribing with openAI’s #Whisper technology? I tried it on myself but now want to try it on something more substantial. anybody want to share experiences?

I'm sure everyone who wants to know about this already does but, just in case anyone has, particularly if #blind or #DeafBlind, been looking for a local method of converting speech to text ... Whisper is an ML model from OpenAI which allows doing that. It can be used accessibly with all screen readers on Windows. Obviously, this is great for those of us with impaired hearing, it is certainly far more accurate than any of the speech to text programs I've seen, needs no training, and can handle background noise quite well. The audio duration limits are set by your hard drive space and the amount of time you're willing to put into transcription, I've transcribed several hours of audio without difficulty, it just takes time. It's available on Windows using github.com/Softcatala/whisper-… which just seems to need python. A GPU makes it faster, but it's usable on an I5 CPU. The model is also available online at freesubtitles.ai though that requires payment or waiting for long periods to transcribe limited amounts of audio. Thanks to @Bryn@mindly.social for the pointer at whisper-ctranslate2. #whisper #SpeechToText

Been dabbling a bit with the amazing #whisper port to C++ coded by @ggerganov. Very good results in both Dutch and English, amazing to see how fast progress has been made on speech to text in the last few years. For those who have got it running and are looking for some tooling: i wrote some Python terminal wrappers for easy use (including converting media using ffmpeg) and converting the SRT files to other formats: github.com/hay/audio2text/