Yeah, whisper is the closest thing we have, but even it requires more processing power than is present in most of these edge devices in order to feel smooth. I've started a voice interface project on a Raspberry Pi 4, and it takes about 3 seconds to produce a result. That's impressive, but not fast enough for Alexa. From what I gather a Pi 5 can do it in 1.5 seconds, which is closer, so I suspect it's only a... - Source: Hacker News / 5 months ago

You can study CTC in isolation, ignoring all the HMM background. That is how CTC was also originally introduced, by mostly ignoring any of the existing HMM literature. So e.g. Look at the original CTC paper. But I think the distill.pub article (https://distill.pub/2017/ctc/) is also good. For studying HMMs, any speech recognition lecture should cover that. We teach that at RWTH Aachen University but I don't think... - Source: Hacker News / 8 months ago

I also tried Kaldi but the build process was too much for my tiny brain; I've also heard good things about vosk but didn't try that. Source: about 1 year ago

Frameworks as well as toolkits like Kaldi were at first promoted by the research study area, yet nowadays used by both scientists and also market experts, reduced the access obstacle in the advancement of automatic speech recognition systems. Nonetheless, cutting edge methods need big speech data readies to achieve a usable system. Source: over 1 year ago

If you interested in unix-like software design and not yet familiar with kaldi toolkit, you definitely need to check it https://kaldi-asr.org It extended Unix design with archives, control lists and matrices and enabled really flexible unix-like processing. For example, recognition of a dataset looks like this: extract-wav scp:list.scp ark:- | compute-mfcc-feats ark:- ark:- | lattice-decoder-faster final.mdl... - Source: Hacker News / over 1 year ago

Yes! I'm currently using https://espeak.sourceforge.net/, so it isn't especially fun to listen to though. Additionally, since I'm streaming the LLM response, it won't take long to get your reply. Since it does it a chunk at a time, there's occasionally only parts of words that are said momentarily. Also of course depends on what model you use or what the context size is for how long you need to wait. - Source: Hacker News / about 1 month ago

You might try espeak or - for something that looks more feature-rich - festival. Source: over 1 year ago

Hey! I’m mute too and I’ve been wanting to stream. So far I’ve decided on using eSpeak https://espeak.sourceforge.net/, a text-to-speech app for PC that allows commercial usage. You might also be able to find online text-to-speech that allows commercial usage, it just might take awhile to find. Depending on the time of content you make you could also dedicate part of your layout to a spot you could type in and... Source: over 1 year ago

Can someone point to a good open source alternative for vocaloid? I know of Sinsy [0] but I couldn't get it working. Ecantorix [1] is very old and rudimentary (it uses espeak underneath [2]). Searching just now I see OpenUtau [3] but I have no experience with it. Seems crazy there isn't a good FOSS solution for this. [0] http://www.sinsy.jp/ [1] https://github.com/divVerent/ecantorix [2]... - Source: Hacker News / over 1 year ago

The closest that I know of is espeak, https://espeak.sourceforge.net/ . It certainly doesn't cover all of the IPA though. Source: over 1 year ago