A demo video in which an 'AI narrator' explains the developer's ecology in real time like an animal documentary program has become a hot topic



Software developer

Charlie Holtz announced ``a system that generates narration audio about one's own ecology in real time using the voice of famous narrator David Attenborough .'' In fact, a demo video that uses Attenborough's voice to create a narration similar to an animal show has been released and has become a hot topic.



Unauthorized “David Attenborough” AI clone narrates developer's life, goes viral | Ars Technica
https://arstechnica.com/information-technology/2023/11/unauthorized-david-attenborough-ai-clone-narrates-developers-life-goes-viral/

The system created by Mr. Holtz that uses Mr. Attenborough's voice to generate narration sounds similar to animal programs in real time uses OpenAI's GPT-4 Vision (GPT-4V), which generates text from images, and ``AI'' from audio samples. It combines ElevenLabs ' technology to generate 'clone voices'.

In order to reproduce Mr. Attenborough's animal program-like explanations and tone with GPT-4V, Mr. Holtz created an API with special prompts. Images captured every 5 seconds by a webcam are fed to GPT-4V via API to generate animal TV-style narration, and the text is then processed using ElevenLabs' AI voice profile trained on Attenborough's voice samples. He said he was having it read out loud.

Mr. Holtz actually uses this system to post a video on X (formerly Twitter) in which Mr. Attenborough's voice narrates ``his own ecology.''

Mr. Holtz setting up the system. The webcam takes a photo of Mr. Holtz every 5 seconds, and the narration is generated based on the image.



When the system was activated, a voiceover began to flow: ``Here is an amazing specimen of Homo sapiens with silver round glasses and long, curly hair.''



Furthermore, Mr. Holtz, who was told something outrageous, can't help but laugh, saying, ``He is wearing something like blue cloth, but I can only think of this as a kind of courtship behavior.''



The narration also mentioned the background, which appeared to be a cafe, and made a surprisingly astute point: ``The background suggests a protected habitat, perhaps a communal feeding or watering hole.''



Mr. Holtz drinks a drink showing off the light blue cup.



'Oh, we're observing sophisticated Homo sapiens engaging in an important ritual of hydration in their natural environment. 'He chooses a small cylindrical container filled with H2O and deftly tilts it toward the opening. What grace and poise.' Holtz explained that he had a drink.



Mr. Holtz has published the code he created to build this system on GitHub.

GitHub - cbh123/narrator: David Attenborough narrates your life
https://github.com/cbh123/narrator



Related Posts:

in Software,   Web Service,   Video, Posted by log1h_ik