API Reference

Data structures

Data structures used to represent speech segments.

segment.Segment(onset, offset)

Speech segment.

Decoding

Functions and classes used to perform SAD on a waveform.

decode.decode(x, sr[, min_speech_dur, ...])	Perform speech activity detection an audio signal.
decode.DecodingError	Error segmenting file.

HTK wrappers

Functions wrapping HTK command line tools.

htk.hvite(wav_path, config, working_dir)	Perform Viterbi decoding for WAV file.
htk.write_hmmdefs(old_hmmdefs_path, ...[, ...])	Modify an HTK hmmdefs file in which speech model acoustic likelihoods are scaled by speech_scale_factor.
htk.HViteConfig(slf_path, hmmdefs_path, ...)	HVite decoding configuration
htk.HTKError	Call to HTK command line tool failed.

IO

Functions for reading/writing segmentatins to files.

io.load_audacity_label_file(fpath[, ...])	Load speech segments from Audacity label file.
io.write_audacity_label_file(fpath, segs[, ...])	Write speech segments to Audacity label file.
io.load_htk_label_file(fpath[, ...])	Load speech segments from HTK label file.
io.write_htk_label_file(fpath, segs[, ...])	Write speech segments to HTK label file.
io.load_rttm_file(fpath)	Load speech segments from Rich Transcription Time Marked (RTTM) file.
io.write_rttm_file(rttm_path, segs, file_id)	Write speech segments to Rich Transcription Time Marked (RTTM) file.
io.write_textgrid_file(fpath, segs[, tier, ...])	Write speech segments to Praat TextGrid file.

Utilities

utils.add_dataclass_slots(cls)	Add __slots__ to a data class.
utils.clip(x, lb, ub)	Clip x to interval [lb, ub].
utils.resample(x, orig_sr, new_sr)	Resample audio from orig_sr to new_sr Hz.
utils.which(program[, search_dirs])	Returns path to excutable program.