Search in a book with speech -


i trying build program find page/sentence in book read microphone. have book's text , audio content. user start reading random page , program supposed synch user , show section of book being read. might seem useless program please bear me..

would approach similar shazam-like programs work? not sure how effective algorithms speech. also, speaker different , might have accent , different speeds read.

another approach converting speech text , searching text in book. problem language of book rare 1 there no language model available. in addition, script not use latin characters makes programming difficult (for me @ least).

is there solutions can recommend? extracting features audio file , comparing "real-time" extracted features (from microphone) work? features?

any implementation/code can start with? language ok prefer c.

you need use speech recognizer.

  1. create language model directly book text. make recognition of book reading accurate, both original reading , reading user.

  2. use language model recognize book , assign timestamps words or use more advanced algorithm perform text audio alignment.

  3. recognize user's speech book-specific language model , use recognized text display position in book.

you can use cmusphinx mentioned tasks.


Comments

Popular posts from this blog

How to remove text and logo OR add Overflow on Android ActionBar using AppCompat on API 8? -

html - How to style widget with post count different than without post count -

url rewriting - How to redirect a http POST with urlrewritefilter -