Based on LPKUnpacker
- Set up the model assets with *.moc3, *.model3.json "sounds" and "motions" directories from .lpk file
- Convert .ogg audios to single-channeled .wav files
- Recount the values of "CurveCount", "TotalPointCount" and "TotalSegmentCount" in *.motion3.json
- Link hit areas with motion group names
- Ffmpeg is needed either in the same dir with main.py or in the system PATH