54696d21 / audiobookForcedAlignment

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Audiobook forced aligner

process:

  1. make whole audiobook to one file mp3cat *.mp3 -o "out.mp3" (must be built from mp3cat.go)
  2. convert book to 16khz ./convertMono.sh out.mp3
  3. ebook-convert mybook.epub mybook.txt OR pdftotext mybook.pdf out.txt
  4. manually delete chapter overview and copyright information from top of the book (can be improved with some NLP in the future)
  5. ./forcedAlign.sh in.wav in.txt -> writes out.srt

to be continued ...

About


Languages

Language:Go 97.3%Language:Shell 2.7%