Improved my text-to-phoneme-to-speech editor, including direct phoneme manipulation, whispering, pig Latin, phoneme substitutions (accents, dialects, and impediments), and gibberish generation using a Markov chain pic.twitter.com/mEymTJsVUT twitter.com/concretescifi/…

In normal circumstances, accurate lip reading is difficult, and it becomes even more difficult and unreliable in loud environments. Yet AI is advancing rapidly, this is a short video demonstrating the translation of lip movements into speech. pic.twitter.com/TzMXDMjQUl