The splitspeech module segments audio feature vectors into speaker clusters.
| Mode | Input | Output | Description |
|---|---|---|---|
_
|
w1, w2, nf
|
w3 | Accepts two word streams, one a sequence of acoustic change points, and the other the classifications from the audiopreproc module, and produces speaker clusters. |
Example:
w3 <- splitspeech3(_, ws:w2, wc:w1, nf)
| DiagCov |
| FixTime |
| MaxNumSpeakers |
| MergeThresh |
| MinNumSpeakers |
|
|