elpis
elpis copied to clipboard
CoEDL
🙊 software for creating speech recognition models.
Resolves #191, #193. This implementation doesn't give the user any choice as to whether to match the file name of the corresponding `.eaf` file or to just get it from...
Updating to react scripts v4.0.3 brings an error in console "The development server has disconnected." Was working previously with v3.4.0
Elpis currently expects annotations to correlate to beginning of media. However, ELAN files can have offset media start times which would result in mis-aligned training. When reading the annotation data...
The training set currently requires matching audio and ELAN file names. For many language doc projects, the filenames differ and renaming could be a large effort. We should match on...
eg, the make_wordlist.py file has a usage note which appears dated
If annotation data contains overlapping or reversed start stop times, we should notify the user so they can correct the data. Overlapping or reversed times can result in "bad segments...
This is doable with pynini and the second approach listed in this paper: The user would have to manually expand the readings for ~300 targeted numbers, and then it can...