TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Getting Deep Speech to Work in Mandarin

58 pointsby kornishover 9 years ago

4 comments

weinzierlover 9 years ago
The amazing part is that their system seems to be adaptable to any language with a minimum of human effort.<p><pre><code> &gt; One of the reasons deep learning has been so valuable is that it has converted &gt; researcher time spent on hand engineering features to computer time spent on &gt; training networks. [...] &gt; We can now train a model on 10,000 hours of speech in around 100 hours on a &gt; single 8 GPU node. That much data seems to be sufficient to push the state of the &gt; art on other languages. There are currently about 13 languages with more than one &gt; hundred million speakers. Therefore we could produce a near state-of-the-art &gt; speech recognition system for every language with greater than one hundred &gt; million users in about 60 days on a single node.</code></pre>
评论 #11109237 未加载
larakernsover 9 years ago
I&#x27;m surprised it doesn&#x27;t use Character Aware Neural Language Models (CNN -&gt; LSTM RNN) but instead a layered RNN. Interesting!
EliRiversover 9 years ago
Facebook disallows some images, based on the personal standards of whoever happens to be in charge of image disallowing that day. Google controls what you see based on your own past, limiting your exposure to opinions you might not like. Companies comply with oppressive government requests for control and surveillance.<p>If we surrender our ability to communicate with people speaking in foreign languages in this fashion, we will literally become unable to talk about things that we &quot;shouldn&#x27;t&quot;, and everything we do talk about will be on permanent record and monitored in real-time for dissent and to target adverts at us.
romanivover 9 years ago
I keep reading about these algorithms that are &quot;better than humans&quot;. Perfect image recognition, perfect speech recognition, parsing plain text-queries and answering questions, etc, etc. So where are the practical implementations?<p>All the speech recognition engines I&#x27;ve interacted with so far were awful. Not just bad, awful.<p><i>&gt;Collecting such data sets could be very difficult and prohibitively expensive.</i><p>Uh, movie subtitles?
评论 #11107045 未加载
评论 #11106113 未加载
评论 #11107869 未加载
评论 #11108124 未加载