Live data from Hacker News

Firefox Voice

voice.mozilla.org

21–30 of 166 posts

Re: Firefox Voice

#21
I'm sorry, but cloud based speech recognition in itself would already be a red flag, even if Mozilla was doing it in-house. Outsourcing it to Google though? I feel like a company as ostensibly privacy-focused as Mozilla should really know better by now...

Re: Firefox Voice

#22
I actually tried this with Siri while cooking yesterday. It's not there yet but I asked "Hey, Siri ... read me the synopsis of the movie Adam's Rib" and Siri proceeded to read a short synopsis of that movie. It worked on another but had to make me choose one of 7. It failed on the 3 try where I tried another movie, it gave me selections, when I picked one "read me the first one" it just repeated the title instead of telling me the synopsis.

Re: Firefox Voice

#23
post #12

I've been working on this exact thing for Chrome the last 3 years: https://www.lipsurf.com Anyone can make an open source plugin for it to do anything with voice ( https://github.com/LipSurf/plugins ) I've wanted to port it to Firefox, but the HTML5 SpeechRecognition API ( https://developer.mozilla.org/en-US/docs/Web/API/SpeechRecog... ) is still not available. Why not just make the API available and leave this in ad…

I built something like this 7 years ago, it's called Hands Free for Chrome, a now languishing project that I lost interest in a long time ago, unfortunately. It made the top 10 of HN back then though! My site's design is not nearly as nice as yours.

https://www.handsfreechrome.com/

I just didn't get enough users or support to really care about it. But I wish you the best. It was an exciting thing to build and using it always felt futuristic to me.

This is just so fascinating though. It's like seeing what could have been if I had been a better developer and found the dedication to really stick to the project in the longterm.

Edit: I see we had the exact same idea! Your "tag" is my "map." Love it. One big difference is that mine was just a free project. I'd be super interested to know how many users you've got. I never had more than ~1100. From looking through your website mine was a much less intensive project. (Oh, CWS says 4000+ for you....wow, wonder how many are paid.)

Edit2: Looking over your update history is almost nostalgic. "Fixed issue with overlapping commands -- delaying commands that are partial matches of other commands." Had to do the exact same thing!

Edit3: We have so many overlapping command names that I wonder if you took inspiration from my project, almost. Either that or it's just a case of convergent evolution.

Edit4: Suggestion for dictation: a way to alternate between a special character and actually writing the word. Doesn't look like there's a way to do ^ vs carrot or & vs ampersand. Something like "Enter special character caret". Maybe you already have a plugin for this though, idk.

Edit5: God, this is so well architected! plugins and contexts are just fantastic ideas for this domain. Click by voice using hidden search-for-text is also a perfect solution to that problem. I wonder if this could be made more intelligent, i.e. "Click Submit in the sidebar on the left"-- challenging though.

Edit6: Wow, just noticed someone else built something called "Handsfree for Web" somewhere along the way and theirs is ALSO way better than what I had built. Geez. Starting to feel bad about my awful website.

Re: Firefox Voice

#27

I'm sorry, but cloud based speech recognition in itself would already be a red flag, even if Mozilla was doing it in-house. Outsourcing it to Google though? I feel like a company as ostensibly privacy-focused as Mozilla should really know better by now...

Yeah, Mozilla is not what it used to be. I guess it's just wishful thinking that makes it so easy to forget.

Re: Firefox Voice

#28
Google-worries aside, judging from the preview it's pretty slow. I'm not a super-fast typer, but these delays sure look like something that would discourage me from actually using it. Maybe it's not even that it's slow, just that the delays are super-obvious somehow, all these disruptive animations and such.

Re: Firefox Voice

#29

I'm sorry, but cloud based speech recognition in itself would already be a red flag, even if Mozilla was doing it in-house. Outsourcing it to Google though? I feel like a company as ostensibly privacy-focused as Mozilla should really know better by now...

They’re building their own open voice platform. Google speech to text is presumably for testing.

> Note: In the future, we expect to enable Mozilla’s own technology for Speech-to-Text which enables us to stop using Google’s Speech-to-Text engine.

edit: s/texting/testing

Post reply on HN