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

DeepSpeech: Scaling up end-to-end speech recognition

arxiv.org

Post reply on HN
Print view View original

21–23 of 23 posts

Re: DeepSpeech: Scaling up end-to-end speech recognition

#21

Post by ogrisel »

pesenti wrote: post #11
cbcase wrote: post #9

Earlier quoted context omitted.

Hi Jerome, those are great results! We got an email this morning from someone else on the Watson team pointing out that we didn't include the latest IBM number -- we'll be sure to update the results in the next version of the paper (three cheers for arXiv). Of course, we openly say in the paper that we don't have the best result on easy subset of Hub5'00 (we had it as 11.5%). We're more interested in advancing the st…

The team is already working on seeing what we get with CH. We'll let you know where we land. But your results are definitely impressive. We love to see new published innovation in the field. Kudos to the team!

What is the average and standard deviation of the performance level on this dataset?

Re: DeepSpeech: Scaling up end-to-end speech recognition

#22

Post by dingdingdang »

cbcase wrote: post #18
dingdingdang wrote: post #14

Earlier quoted context omitted.

Are any of these systems open source?

Both Kaldi[1] and CMU Sphinx[2] are high-quality open source speech systems. I know for a fact that Kaldi includes support for DNN acoustic models (I'm less familiar with Sphinx). [1] http://kaldi.sourceforge.net/ [2] http://cmusphinx.sourceforge.net/

Thanks, appreciated, but my dear lord, without a PhD in AI systems these things are a bit beyond what most users, me included, would casually play around with. Be great if this tech made it into Dragon Naturally Speaking-like end product to use privately.

Re: DeepSpeech: Scaling up end-to-end speech recognition

#23

Post by pesenti »

pesenti wrote: post #11
cbcase wrote: post #9

Earlier quoted context omitted.

Hi Jerome, those are great results! We got an email this morning from someone else on the Watson team pointing out that we didn't include the latest IBM number -- we'll be sure to update the results in the next version of the paper (three cheers for arXiv). Of course, we openly say in the paper that we don't have the best result on easy subset of Hub5'00 (we had it as 11.5%). We're more interested in advancing the st…

The team is already working on seeing what we get with CH. We'll let you know where we land. But your results are definitely impressive. We love to see new published innovation in the field. Kudos to the team!

For CH we get 19.1% for a combined rate of 14.75% - this is using 300 hours of training data.
Post reply on HN