vadimsc

7 ms·

1.
by vadimsc 2mo ago
Thank you for the warm words! Model already generalizes on out-of-distribution people, so you actually don't need to train a model for each person
2.
by vadimsc 2mo ago
One cool fact: with silent speech, you can “speak” not only while exhaling, but also while inhaling
3.
by vadimsc 2mo ago
Voice is still a very underexplored modality. Right now, it’s used mostly for dictation, but it could also be used for editing. The funny part is that the tongue itself could act as a cursor or perform gestures, which sets this approach apa
4.
by vadimsc 2mo ago
It won’t necessarily be adhesive. It could be attached mechanically - we’re still exploring the form factor. The idea is that you won’t need cubicles anymore. Your hands will always be free because you won’t need a keyboard or mouse. You’ll
5.
by vadimsc 2mo ago
Exactly - a universal input device. It could help not only people with vocal cord problems, but also essentially anyone who is paralyzed from the neck down.
6.
by vadimsc 2mo ago
<3
7.
by vadimsc 2mo ago
When I was collecting data, I spoke silently for more than two hours. After the session, I tried to use Wispr Flow but forgot that I was still speaking silently, so it didn’t work. When I finally said it out loud, it felt like such a hassle
8.
by vadimsc 2mo ago
Tongue movement detection cannot track only six phonemes: - b - p - m - f - v - h That’s a remarkably small number compared with other methods. English could easily be optimized to achieve a sub-5% WER with a very small model.
9.
We fixed the News at AGI House hackathon(twitter.com)
2 points by vadimsc 3y ago|1 comments
10.
by vadimsc 3y ago
We developed an app that sets common knowledge about the event and highlight contradictions and divergences of opinion. For example this is what we got about Sam Altman getting fired: Commonalities: OpenAI's board fired CEO Sam Altman