Apple announces 'Keyframer,' an AI tool that animates still images simply by entering prompts
On February 8, 2024, Apple's research team announced Keyframer , an AI tool that uses large-scale language models to animate still images.
Keyframer: Empowering Animation Design using Large Language Models
Apple researchers unveil 'Keyframer': An AI tool that animates still images using LLMs | VentureBeat
https://venturebeat.com/ai/apple-researchers-unveil-keyframer-an-ai-tool-that-animates-still-images-using-llms/
Apple is developing an AI tool for animating images using text prompts - The Verge
Keyframer, which uses OpenAI's large-scale language model GPT-4 as its base model, uses user-uploaded SVG file images to generate CSS code to implement user-entered text prompts.
Users can create animations from still images in Keyframer by simply following three steps: upload an image, enter a text prompt, and click 'Generate.'
In a demo, Apple shows how the background changes when you type a prompt like 'Twinkle the stars.'
Users can also freely adjust properties such as the color code used in the animation and the animation duration. Keyframer automatically converts the changes you make into CSS, so no complex coding knowledge is required. You can also edit the CSS code directly.
'Keyframer supports further exploration and refinement of animation through a combination of prompts from generated output and direct editing,' Apple explains.
One professional motion designer who participated in Apple's Keyframer test said, 'I'm worried that my job will be replaced by Keyframer. The potential of Keyframer is so great. We intend to use Keyframer as one of our animation tools.'
However, Apple pointed out that the Keyframer test was conducted on a small sample of just 13 people, and that the tool was only available to test on two pre-selected images in SVG file format.
Apple also clarified that Keyframer is focused on creating web-based animations, such as sequence import, data visualization, and animated transitions, and is not currently capable of creating the complex animations seen in movies and games.
Apple has announced its focus on generative AI , and in February 2024 released the image editing AI model 'MGIE.'
Apple releases MGIE, an image editing AI model that edits photos simply by giving verbal instructions, and a demo has been released, so I tried it out - GIGAZINE
Related Posts:
