AI Listened to People's Voices. Then It Generated Their Faces.
#1
This is arch-creepiness to me. AI is getting very, Very sophisticated and has the potential to be fully intrusive into our lives and our way of life as well.

Interesting article this:


Keyboard Warrior




Article Wrote:Link to Original Article

 
AI Listened to People's Voices. Then It Generated Their Faces.
Mindy Weisberger, Senior Writer
4-5 minutes

[Image: aHR0cDovL3d3dy5saXZlc2NpZW5jZS5jb20vaW1h...JvbS1mYWNl]
The algorithm approximated faces based on gender, ethnicity and age, rather than individual characteristics.
Credit: Oh et. al.

Have you ever constructed a mental image of a person you've never seen, based solely on their voice? Artificial intelligence (AI) can now do that, generating a digital image of a person's face using only a brief audio clip for reference.

Named Speech2Face, the neural network — a computer that "thinks" in a manner similar to the human brain — was trained by scientists on millions of educational videos from the internet that showed over 100,000 different people talking.

From this dataset, Speech2Face learned associations between vocal cues and certain physical features in a human face, researchers wrote in a new study. The AI then used an audio clip to model a photorealistic face matching the voice. [5 Intriguing Uses for Artificial Intelligence (That Aren't Killer Robots)]

The findings were published online May 23 in the preprint jounral arXiv and have not been peer-reviewed.

Thankfully, AI doesn't (yet) know exactly what a specific individual looks like based on their voice alone. The neural network recognized certain markers in speech that pointed to gender, age and ethnicity, features that are shared by many people, the study authors reported.

"As such, the model will only produce average-looking faces," the scientists wrote. "It will not produce images of specific individuals."

AI has already shown that it can produce uncannily accurate human faces, though its interpretations of cats are frankly a little terrifying.

The faces generated by Speech2Face — all facing front and with neutral expressions — didn't precisely match the people behind the voices. But the images did usually capture the correct age ranges, ethnicities and genders of the individuals, according to the study.

However, the algorithm's interpretations were far from perfect. Speech2Face demonstrated "mixed performance" when confronted with language variations. For example, when the AI listened to an audio clip of an Asian man speaking Chinese, the program produced an image of an Asian face. However, when the same man spoke in English in a different audio clip, the AI generated the face of a white man, the scientists reported.

The algorithm also showed gender bias, associating low-pitched voices with male faces and high-pitched voices with female faces. And because the training dataset represents only educational videos from YouTube, it "does not represent equally the entire world population," the researchers wrote.

Another concern about this video dataset arose when a person who had appeared in a YouTube video was surprised to learn that his likeness had been incorporated into the study, Slate reported. Nick Sullivan, head of cryptography with the internet security company Cloudflare in San Francisco, unexpectedly spotted his face as one of the examples used to train Speech2Face (and which the algorithm had reproduced rather approximately).

Sullivan hadn't consented to appear in the study, but the YouTube videos in this dataset are widely considered to be available for researchers to use without acquiring additional permissions, according to Slate.

Originally published on
Live Science.
One should have an open mind; open enough that things get in, but not so open that everything falls out
 
A democracy which makes or even effectively prepares for modern, scientific war must necessarily cease to be democratic. No country can be really well prepared for modern war unless it is governed by a tyrant, at the head of a highly trained and perfectly obedient bureaucracy
Huxley
 
The world will not be destroyed by those who do evil, but by those who watch them without doing anything  
Einstein
 
Its no wonder truth is stranger than fiction.
Fiction has to make sense
Mark Twain
Reply
#2
The nasal sound in your voice will have a significant impact on computer generated appearance.  Lisps and other speech impediments also.
.
I don't know how regional/national accents will affect this.
.
Yeah, it is a bit creepy.
[-] The following 1 user Likes MaryTN's post:
  • Zedta
Reply
#3
Most Americans can identify a black person by the time they finish saying  "hello" upon answering a telephone. The problem really isn't with technology but with the sociolinguistics of English and the particular character of our societies.
[-] The following 1 user Likes Cyriacus's post:
  • Zedta
Reply
#4
(06-13-2019, 05:18 PM)Zedta Wrote:
Article Wrote:The algorithm also showed gender bias, associating low-pitched voices with male faces and high-pitched voices with female faces.

And they say we're the science deniers.
[-] The following 1 user Likes Paul's post:
  • Zedta
Reply




Users browsing this thread: 1 Guest(s)