Live-caption glasses let deaf people read conversations [video]

Post by BeFlatXIII » Wed, Mar 15, 2023, 2:03 PM UTC

Kalium wrote: post #73

lazyeye wrote: post #64

Earlier quoted context omitted.

Completely agree. There seems to be such a gap in the market for privacy-first hardware but very few companies are doing anything.

The key question is always a painfully simple one: how much are people willing to pay as a privacy-first premium? The answers to this determine everything. Treating privacy-first as the moral and ethical default we should expect everyone to start from is a wonderful idea, rooted in compassion, kindness, and a foundational respect for human rights. It has also been an abject failure to date. We should not expect the f…

Post by danscarfe » Wed, Mar 15, 2023, 11:32 PM UTC

quantumquetzal wrote: post #141

ape4 wrote: post #32

Earlier quoted context omitted.

It would be nice if it put the captions over the speaker (in a speech bubble?)

There’s a significant amount of evidence pointing to the fact that “anchored” captions (a la speech bubbles, like a comic book) is a preferred way to render captions, assuming the bubbles are at the same focal distance as the speaker. This is solvable with “true” AR, but getting that compute into the lightweight form factor of the original video is an unsolved problem, and is a ways off.

Post by danscarfe » Wed, Mar 15, 2023, 11:33 PM UTC

flangola7 wrote: post #145

quantumquetzal wrote: post #15

Background: researched this space for a graduate degree. There are a few issues that are unanswered by this video (which isn't intended to be a technical deep dive, but I don't see any related links in the video description): 1. How do these glasses handle multiple simultaneous speakers? Based on the display I saw, it shows the speakers' words sequentially, which starts to fall apart in real-world environments, espec…

The Android voice recording app can already write a transcript in real time that indicates different speakers.

Post by midoridensha » Thu, Mar 16, 2023, 1:53 AM UTC

elliekelly wrote: post #123

generalizations wrote: post #111

Earlier quoted context omitted.

That's fair. It's easy to read hostility where there isn't any. > ASL isn’t a “foreign” language. It’s an American language. Elsewhere in this thread you mentioned that "ASL is a complete and distinct natural language in its own right," and that's the sense I was trying to convey by calling it a "foreign" language. It's as hard to acquire as a foreign language, and carries the same benefits. And it fit the analogy. I…

Do you realize that “learn to lip read” isn’t feasible for a lot of people? That’s great that lip reading worked for the one Deaf person you know. But Deaf people, like the population at large, are all different people. It’s cruel and ridiculous to suggest all Deaf people should just learn to lip read. I tried to explain the cruel history behind your callous remark. I tried to give you an opportunity to empathize wit…

Post by rapjr9 » Thu, Mar 16, 2023, 4:25 PM UTC

kajecounterhack wrote: post #67

roughly wrote: post #65

Earlier quoted context omitted.

I imagine it’d substantially increase the compute load, but I’d be curious if you could use multiple microphones and beam forming to separate out the streams of speech and feed them to the TTS algorithm independently.

Single-mic source separation is possible in an unsupervised manner today that could probably work better than beamforming both compute-wise and with regard to implementation difficulty (you'd just need a lot of recordings to represent the space of sounds you want to separate).