CTC segmentation

This is an unofficial republication of lumaku/ctc-segmentation maintained by the ESPnet project. It is not affiliated with, nor endorsed by, the original authors. It exists only because upstream's released 1.7.4 builds against the NumPy 1.x ABI and so fails to import under NumPy 2 - a problem upstream has already fixed on master but has not released. The import name is unchanged. See FORK_NOTICE.md.

CTC segmentation is used to align utterances within audio files. It can be combined with CTC-based ASR models. This package includes the core functions.