Papers of the day   All papers

Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis

Comments

Binni Shah: Real-Time-Voice-Cloning : Clone a voice in 5 seconds to generate arbitrary speech in real-time : https://github.com/CorentinJ/Real-Time-Voice-Cloning Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis : https://arxiv.org/pdf/1806.04558.pdf (pdf) Demo : https://www.youtube.com/watch?v=-O_hYhToKoA https://t.co/iRLTfA2tZW

22 replies, 1384 likes


Károly Zsolnai-Fehér: This AI Clones Your Voice After Listening for 5 Seconds ▶️Full video (ours): https://www.youtube.com/watch?v=0sR1rU3gLzQ 📜Source paper: https://arxiv.org/abs/1806.04558 #ai #deeplearning #science #twominutepapers https://t.co/BZOxdNgi4Z

6 replies, 93 likes


Károly Zsolnai-Fehér: This AI Clones Your Voice After Listening for 5 Seconds ▶️ Full video (ours): https://youtube.com/watch?v=0sR1rU3gLzQ 📜 Source paper: https://arxiv.org/abs/1806.04558 #ai #deeplearning #science #twominutepapers https://t.co/PDTUbNW82M

1 replies, 68 likes


ML Review: "Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis" TTS is able to generate speech audio in the voice of many different speakers, including those unseen during training. Samples https://google.github.io/tacotron/publications/speaker_adaptation/index.html Arxiv https://arxiv.org/abs/1806.04558 https://t.co/IjSXuJMAkc

0 replies, 62 likes


Empereur de la Lune: @MohamedGhilan For the white paper this is about see here: https://arxiv.org/pdf/1806.04558.pdf For the code to clone a voice see here: https://github.com/CorentinJ/Real-Time-Voice-Cloning For examples of cloned voices see here: https://google.github.io/tacotron/publications/speaker_adaptation/

1 replies, 54 likes


Hurricane Jerry: Security awareness training classes are about to get much more involved. We are going to need to start using mutual two factor authentication when we talk on the phone or on video conferences soon.

1 replies, 19 likes


Alison B Lowndes ✿: Play around with Voice cloning! Great article by @GeorgeSeif94 on @Google's https://medium.com/p/you-can-now-speak-using-someone-elses-voice-with-deep-learning-8be24368fa2b Paper: https://arxiv.org/pdf/1806.04558.pdf #deeplearning #NLP #TTS

0 replies, 7 likes


𝖆𝖑𝖎𝖘𝖔𝖓 | sextech: Easy access for deepfakes, deepnudes, and now deep audio

2 replies, 4 likes


Tyrone E. Wilson: 👀😳😳😳

2 replies, 4 likes


Chey Cobb, cissp, lsmft, iirtplayn: Shhhhiiii... @thespybrief

1 replies, 3 likes


Alexey Vidanov: Very impressive tool 😱 to clone voices. Need to test it 👷‍♂️. #VoiceFirst

0 replies, 3 likes


Matteo: "deep fake audio" by @CorentinJemine at @resembleai . Some may say its not as convincing as @dessa 's "joe rogan" but this is open-source &created very quickly. The other stuff from resemble ai, also worth a look. https://github.com/resemble-ai/Resemblyzer AND https://github.com/CorentinJ/Real-Time-Voice-Cloning #opsec #ai

0 replies, 3 likes


Renato Candido: A neural network-based system for text-to-speech synthesis that clones voices after listening for 5 seconds https://www.youtube.com/watch?v=0sR1rU3gLzQ https://arxiv.org/abs/1806.04558

0 replies, 2 likes


Prasanna Srikhanta: 👇🤯it's amazing how quickly these techniques are evolving. the next frontier will be adding emotional tone to generated voices.

0 replies, 2 likes


Desert Rose Lee🔮🌵🔮: I have plans.

0 replies, 1 likes


Javier Lombardi: @GrupoIngSocial #impersonation

0 replies, 1 likes


Abdessalem Hammami: "Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis" Samples: https://google.github.io/tacotron/publications/speaker_adaptation/index.html Arxiv: https://arxiv.org/abs/1806.04558 Code: https://github.com/Swall0w/papers/issues/496 #AI #IntelligenceArtificielle #MachineLearning #DeepLearning #neuralnetworks #NLProc https://t.co/OAHncx770J

0 replies, 1 likes


Mr. Bill: MY VOICE IS MY PASSPORT VERIFY ME 😮

0 replies, 1 likes


Rory Byrne: I saw reference to this sort of tech years ago being used by military units (Israel I think it was) to gain various advantages - vectoring enemy aircraft/people to wrong locations, creating confusion etc. Interesting to see it now available in public and the effects it will have.

1 replies, 0 likes


Content

Found on Sep 18 2019 at https://arxiv.org/pdf/1806.04558.pdf

PDF content of a computer science paper: Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis