Hey guys & gals, and I’m a founder of just created company https://voixware.com and I’m going to publish the first commercial product - a real-time subtitler of computer sound, speech to subs - https://speech2subs.com. I have profound hearing loss and created this app firstly for my own needs.
If you would like to get early access to this program, or just feel interest - please submit this very short form https://forms.gle/WLpabZ2hm63DUfWTA
Lets me present myself.
I started to lose my hearing since childhood (after a traumatic brain injury when I was 3 y.o.). So it was challenging enough to realize myself in real life, I think the readers of this thread know it very well.
Fortunately, since childhood I like computers. My mother (she worked in a bookshop) purchased me books about computers and programming - and I was very impressed. We had no computer at that time, it was 198x years and I was living in the Soviet Union (Russia). In 1989 my father (he was an engineer) built ZX Spectrum 80 computer - he soldered it from contraband electronic details purchased in Odesa (Ukraine). I was 6-7 y.o. and I was really amazed to work with a real computer. Of course, I played the games, but also I learned how to program.
After that time my family purchased a couple of other computers - one soviet in 1994 and then IBM PC compatible in 1996. I was continuing play games and programming with help of paper books. Then I was studied in the university on the Faculty of Applied Mathematics and Informatics. Then I worked in a few IT companies, web-startups, e-commerce projects. Summary at this moment I have almost 20 years of commercial working experience, including taking a part in some startups from a scratch.
Some time ago I was relocated to the US and started work with an American company. It was really hard for me to found a job in the US - with my hearing loss I can’t recognize a word by phone, but no other way exists to apply to a job in the US. My native language was Russian and my brain was trained to recognize Russian speech - not very certain with so hard hearing loss, not in the noise of course, but nevertheless… But for my brain and damaged hearing nervous it’s a problem to chat by phone and recognize foreign languages speech even in real life. So it was a reason I start figuring out how to help myself with call chats.
As a result, I started creating the subtitling application. The idea is pretty simple - my software catches computer sound input and writes the subtitles of speech in real-time. I developed this app for 3 years - and have been using it for my own goals the whole time. With help of this app I was settled in the US (all these rentals, banking, utility calls - I think you know these crazy unlimited chats). And thanks to my app, I was processing the job interviews and successfully got an offer. So this program is proven by my own experience.
Thus I am definitely sure my software is helpful for all hearing-loss communities and improves the quality of life. Phone calls and real-time voice chats are not a problem anymore for people with partial hearing loss, and foreigners with non-excellent listenings skills as well.
I have worked a full-time job until now, and these days I just created my new startup VoixWare - to make my main job the development of this and other voice-related apps.
The speech2subs app is still developing now - I’m adding multi-user possibilities and creating a personal online cabinet, new improvements. I’ll inform the community about releases and I’d like to get feedback about your thoughts about it. I’m planning to open testing for a limited pool of people. Also, I’d like to know the feedback from all the people who read this speech till the rest.
Thanks for reading and don’t hesitate to ask me any questions.
Google Live Transcribe is an Android app.
Speech2subs is creating for macOS but, because of cross-platform architecture, will work also with Windows as well.
When I was testing Chrome browser real-time captioning, it was working only with video.
Yes, I use it for my calls - Zoom, Skype, phone. Of course, as with other speech recognition services, many depend on the quality of speech: clear/indistinct pronunciation, noisy background, accent. Some people are recognized fine, some - not good. It’s not a magical wand, but without this kind of soft, I can’t recognize even a phrase)
How does it work with IPhone? Google transcribe is great for Android phones but the corresponding AVA app on iPhone does not capture anything more than a few inches away.
@nelhop I did it for my macOS, also it is compilable for Windows. At the current time, speech2subs doesn’t work with phones because it’s a different architecture.
For real-life I recommend otter.ai for iPhone - I use it in offline situations, and it works fine.
Pro tip for users of any mobile applications for speech recognition: if you have bluetooth HA - disable Bluetooth first! Otherwise otter (or any another app) will work worse (because it tries to capture sound from hearing aids, not phone microphone) and you’ll have less sound volume (because during Bluetooth streaming HA’s microphones volume usually decreases). I do all in a few clicks: disable all Bluetooth on iPhone from quick settings, then start otter ai recognition (I attached this app to menu bar/dock).
I wanted to update the community on the status of the application. I am happy to announce that the application is now available for public use on MacOS. The Windows version is planned for release in 2023.
Since I announced about this app 1.5 years ago, the app development has come a long way from a prototype that only worked on my laptop to a safe Apple-signed distributive for MacOS. In addition, a lot of additional backend parts and extensive research and testing of various external services were processed in the development of the app.
Now the application works even with a loudly environment or with unclear voices, and the accuracy of speech recognition is much higher than it was in the beginning. It is also designed to recognize different languages and has an additional feature of translating foreign language speech to English (or another native language).
If you have a Mac computer and would like to try it out, please visit speech2subs.com. Don’t hesitate to reach out to me in case of any questions.
It’s been a while since my last post, but I’m super excited to share some awesome updates to our program, VoixWare speech2subs.
First off, we’ve added a bunch of new languages for speech recognition, not just English. Plus, we’ve introduced a feature that lets you transcribe speech in two languages at once - the original language and an instant translation into another language. So now, our program is even more useful for non-native speakers and those learning a foreign language.
Another sweet improvement is that you can now choose to listen to the microphone as well, instead of just your computer’s audio output. We’ve also improved speech recognition accuracy, so with microphone listening, you can “listen” to people seated in the room, like during office meetings. And yes, subtitling in two languages at once works in this mode too.
We hope you love these updates as much as we do, and we can’t wait to hear what you think. Keep the feedback coming!