This topic created in 1877 days ago, the information mentioned may be changed or developed.
Satin: Microsoft’s latest AI-powered audio codec for real-time communications
"Satin is already being used for all Teams and Skype two-party calls and will roll out for Teams meetings soon."
Lyra: A New Very Low-Bitrate Codec for Speech Compression
"Duo already uses ML to reduce audio interruptions, and is currently rolling out Lyra to improve audio call quality and reliability on very low bandwidth connections."
链接里面有英文样本（程序未发布）
据说可以实时处理，应该是最近一年网络会议需求催生的的产品
Lyra 3kbps，Satin 6kbps，走 ML 模型，按照我的理解是把音频提炼成字体文件那样再还原
我们的 AI First 公司啥时候也出一个啊
 |
1
no1xsyzy Mar 10, 2021
自动提取为 Vocaloid / UTAU 音源（不是
|
 |
4
mikumkf Mar 10, 2021
类似的方法在国内已经被应用的很广泛的。直观来说是类似于 PCA 而不是和识别相关的成字体，先用 Encoder 成一个编码，传输后再用 Decoder 还原细节。
|
Select Language