Kaldi-compatible online & offline feature extraction with PyTorch, supporting CUDA, batch processing, chunk processing, and autograd - Provide C++ & Python API
Release the GIL in offline feature extration
Support building on Windows.
Support streaming feature extractors.
Move audio data to GPU in a chunk-wise manner to avoid CUDA OOM for long utterances, i.e., those of more than 20 hours long.
Add pickling support.
Add from_dict
and as_dict
to various Options
in kaldifeat
Support using kaldifeat in k2.