3 ms·
2 minute papers has an episode on a Microsoft AI that can clone a voice in 3 seconds. Probably not good enough to break this (4 minutes of voice in the article)
by pixl97 4y ago
2 minute papers has an episode on a Microsoft AI that can clone a voice in 3 seconds. Probably not good enough to break this (4 minutes of voice in the article), but ya voice is not a security feature any longer by any means.
https://youtu.be/F6HSsVIkqIU https://youtu.be/F6HSsVIkqIU
- thangalin 4y agoI've been using Tortoise TTS[1] to narrate the sci-fi novel I'm writing (alpha readers wanted, see profile for contact): $ find audio/ -type f -name "*combined*wav" | sort -n audio/01-xander/xander_combined.wav audio/01-yuna/yuna_combined.wav audio/02-xander/xander_combined.wav audio/02-yuna/yuna_combined.wav audio/03-cassandra/cassandra_combined.wav audio/03-yuna/yuna_combined.wav audio/04-xander/xander_combined.wav audio/04-yuna/yuna_combined.wav audio/05-cassandra/cassandra_combined.wav audio/05-yuna/yuna_combined.wav audio/06-cassandra/cassandra_combined.wav audio/06-yuna/yuna_combined.wav audio/07-xander/xander_combined.wav audio/07-yuna/yuna_combined.wav audio/08-cassandra/cassandra_combined.wav I started the task on March 10. My NVIDIA T1000 GPU (no joke) has been chewing through the prose non-stop. There are 12 chapters written so far, with the computer narrating about a chapter a day. I can't share the audio for legal reasons (celebrity voice snippets plucked from YouTube interviews), but the output is nearly human in quality, though there are numerous glitches in the o̶u̶t̶p̶u̶t̶ matrix. Here's the script that launches Tortoise TTS: for i in $HOME/.../chapter/??-*.txt; do F=$(basename $i); VOICE=$(echo ${F%.*} | cut -c 4-); CHAPTER=$(echo ${F%.*} | cut -c -2); D=$(dirname $i); CLIPS="$D/../audio/$CHAPTER-$VOICE"; mkdir -p $CLIPS; ./scripts/tortoise_tts.py -O $CLIPS -v $VOICE < $i; done [1]: https://git.ecker.tech/mrq/tortoise-tts https://git.ecker.tech/mrq/tortoise-tts
- snerbles 4y agoWhat's your review workflow? Do you pre-generate extra takes? I've played around with doing TTS narration for some webfiction authors, reviewing some of the more wild outputs for Tortoise's occasional demonic weirdness can be quite time consuming.