In this forum you can talk about Anime and Cartoons such as Hunter X Hunter, Dragon Ball, The Seven Deadly Sins, Ben Ten, Scooby Doo, Kuroko's Basketball, Shaman King, Baki Hanma, KENGAN ASHURA, Fullmetal Alchemist, One Piece, Jujutsu Kaisen, K-On, Spirited Away, Negima, Powerpuff Girl, X-men, Dexter's Laboratory, Star Wars, Samurai Jack, The Simpsons, Teen Titans, Avatar: The Last Airbender, SpongeBob, Batman, Rick and Morty, Looney Tunes, Tom and Jerry, Attack on Titan, Neon Genesis Evangelion, Naruto, and any others. Any guest can reply to any topic if not locked. You can also share your opinion or stuff about anime and general everyday topic.
.

Post reply

Warning: this topic has not been posted in for at least 120 days.
Unless you're sure you want to reply, please consider starting a new topic.
Name:
Email:
Subject:
Message icon:

Verification:
One Piece's main protagonist:
Guku's race:
fdgdfgdfg:

shortcuts: hit alt+s to submit/post or alt+p to preview


Topic Summary

Posted by: Yudas
« on: March 02, 2024, 10:58:51 PM »

Alibaba is launching EMO, an artificial intelligence that creates singing videos from selfies, in hopes of taking on the GenAI battle.

The Gen AI battlefield continues to rage. After the launch of OpenAI's Sora, many tech companies have plans to create their own AI builders. One of them is Alibaba, the technology giant from China.

Alibaba's Intelligent Computing Research Unit revealed that preparing to create an "EMO" video based on artificial intelligence. Using just one image, you can create characters that sing, talk, move their faces and show different facial expressions.

EMO's video presentation uses portraits along with audio files that require the characters to speak. Which can make a maximum duration of 1 minute and 30 seconds. The expression of a person's facial expression is accompanied by emotion and the sound that comes out. It's not just about moving your mouth.

As for the character singing function, it can be adjusted according to the language of the song. In addition to quickly jumping to the beat of the music, the AI can also differentiate between English and Korean voices.

EMO is supposed to be short for Emotive Portrait Alive, but in that case, why not EPO? The manipulation of the word "emo (emo)" is a unique feature of its AI: the characters in the videos created by EMO can speak. She can sing, but her rival Sora can't speak.



        "Instead of just creating characters, actors look at each other back and forth. Why not create one where they can talk to each other? So EMO is the future of artificial intelligence that can create virtual worlds. With characters that can interact with each other .”, said the creators of EMO. Some in a lively way."

EMO offers a more advanced video example: a clip of a Japanese woman walking down the street inspired by Sora, but EMO makes the woman sing. Dua Lipa's song "Don't Start Now" is her go-to song.

However, the EMO model is still in the development stage. He has to go through the same moral process that Sora went through. In the sample clips created by EMO, most characters do not show strong emotions, such as being very angry or crying a lot.

Alibaba is committed to becoming the world's leading AI technology company. The goal is to use AI to drive business growth. Create new innovations and improve people's quality of life The previous innovation is the development of a large-scale language model WuDao 2.0 with OpenAI's GPT-3 equivalent; launching the Alibaba AI Academy, an AI training institute; and joint investments with leading universities around the world to develop artificial intelligence technology.

Disclaimer : All pictures and media in this site and forum copyright to their respective owner. If you see any of your media on this site or forum and don't want it here, send a message with the details and the link to the media and we will remove it.