dongyuan4790 2018-12-11 16:34
浏览 32

Google Cloud Speech-To-Text会丢弃FLAC文件块

I had a voice-to-text app working on a server, but I switched servers, and now it's not working any longer. I talk for a minute (for example), it's converted on the server to an accurate flac file, and then sent to the Google Cloud to be converted. When it comes back as text, there are large chunks of text missing. I throw in a "10 second", "20 second" marker just to see where it's dropping the chunks, and I get back something like the below:

"testing to see if this will work at the time I am at 7 seconds and then I'm going to add some more interesting to see if this will work at 22nd at 42nd without any punctuation keep talking and now I am at 1 minute."

This should have a 10, 20, 30ish marker every 10 seconds. So random things dropped all over, but big chunks missing between 20 & 40 and 40 & 60.

Does anyone have any idea where I can start to find out what is the issue? Any error reporting anywhere in the console? Somewhere I can upload the flac manually and get a real time result? I dont know if this is a code, server, or Google issue.

  • 写回答

0条回答 默认 最新

    报告相同问题?

    悬赏问题

    • ¥15 装 pytorch 的时候出了好多问题,遇到这种情况怎么处理?
    • ¥20 IOS游览器某宝手机网页版自动立即购买JavaScript脚本
    • ¥15 手机接入宽带网线,如何释放宽带全部速度
    • ¥30 关于#r语言#的问题:如何对R语言中mfgarch包中构建的garch-midas模型进行样本内长期波动率预测和样本外长期波动率预测
    • ¥15 ETLCloud 处理json多层级问题
    • ¥15 matlab中使用gurobi时报错
    • ¥15 这个主板怎么能扩出一两个sata口
    • ¥15 不是,这到底错哪儿了😭
    • ¥15 2020长安杯与连接网探
    • ¥15 关于#matlab#的问题:在模糊控制器中选出线路信息,在simulink中根据线路信息生成速度时间目标曲线(初速度为20m/s,15秒后减为0的速度时间图像)我想问线路信息是什么