TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: How would you build a text to speech startup?

1 点作者 h3ctic将近 5 年前
Since Google's WaveNet TTS systems shouldn't sound robotic but they still do. Third party TTS engines for Android aren't good and those who did sound good have been bought and shut down. Which markets would you approach? What problem of yours can be solved?

1 comment

PaulHoule将近 5 年前
Imagine you were making a visual novel:<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Visual_novel" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Visual_novel</a><p>you could write one is not voiced, voiced with some exclamations (&quot;Are you sure that&#x27;s safe?&quot; gets replaced with &quot;Yikes!&quot;) voiced in some places and not voiced in others, or fully voiced.<p>In 2020 you need to hire and direct a voice actor because TTS is nowhere near good enough. If there was a good enough TTS you wouldn&#x27;t need to hire the voice actor although you&#x27;d still need to direct the TTS.<p>Two of my favorite speakers are Lester Holt and Tanaka Rie and I&#x27;d like to see a computer voice that approaches that kind of person.