Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

High-performance speech recognition with no supervision at all

ai.facebook.com

Post reply on HN
Print view View original

31–40 of 66 posts

Re: High-performance speech recognition with no supervision at all

#31

Post by joering2 »

local_dev wrote: post #26
superqd wrote: post #23

Earlier quoted context omitted.

I was contacted by their recruiters recently and started getting warmed up on their interviewing, etc. However, this was back in December when they put out ads against Apple's choice to allow users to opt out of tracking. It kind of made it difficult to want to follow through on the interview process after that... so I didn't.

I did the same. I don't think you are alone, at all. FB pays well and works on really interesting things but I just can't justify the moral price of working there.

Thank you for your choice of not working there!

Re: High-performance speech recognition with no supervision at all

#35

Post by pesenti »

skybrian wrote: post #14

How does Facebook use speech recognition in its own products?

Our main usages of speech rec are:

- Understanding for videos for content moderation, ranking and recommendations

- Captioning

- Assistant tech in Portal, Oculus and upcoming glasses

Re: High-performance speech recognition with no supervision at all

#36

Post by garmaine »

savant_penguin wrote: post #9
imedadel wrote: post #6

Facebook's focus on unsupervised machine learning is a huge plus for under-resourced languages. They had a similar article for unsupervised machine translation[1] before, and I can see how it'll open doors for many African languages. [1] https://engineering.fb.com/2018/08/31/ai-research/unsupervis...

Not only that but it's also great for the English models as well. Many models are trained using the libre speech dataset with 960h of labeled audio. What if you trained with 960000 hours instead??

Labeled data is easy to come by in popular language though. Just get an Audible subscription.

Re: High-performance speech recognition with no supervision at all

#38

Post by croes »

pesenti wrote: post #35
skybrian wrote: post #14

How does Facebook use speech recognition in its own products?

Our main usages of speech rec are: - Understanding for videos for content moderation, ranking and recommendations - Captioning - Assistant tech in Portal, Oculus and upcoming glasses

And the not main usage?

Re: High-performance speech recognition with no supervision at all

#39

Post by unityByFreedom »

croes wrote: post #37
unityByFreedom wrote: post #29

That's incredible. Since translation can be done this way too, it may truly be a short time until widespread use of babel-fish tech becomes reality. What a world!

Nope, like autonmous driving the problem will be the last 20%

Just because one overhyped tech is not working out does not mean all new tech will fail.

We already have noise cancellation, good tiny batteries, fast wireless internet, and all the tech to learn spoken languages without supervision!

Autonomous driving has a much higher bar for accuracy.

Re: High-performance speech recognition with no supervision at all

#40

Post by strikelaserclaw »

unityByFreedom wrote: post #29

That's incredible. Since translation can be done this way too, it may truly be a short time until widespread use of babel-fish tech becomes reality. What a world!

Babelfish like technology would be a dream come true for me. At this point in time, google translate speech recognition works well when the speaker speaks a bit slow and uses simple sentences. The translation is the same, works well for simple sentences even though the translations in my language sound weird because it uses the equivalent of old English. There is still a way to go before speech recognition can recognize fast speech with colloquialisms, and probably even a longer way to go before it can translate longer sentences with abstractions into something that is clear and concise in the target language.
Post reply on HN