Image input field to add SVG code and preview the image; GPT prompt section for entering a natural language prompt. Credit: arXiv (2024). DOI: 10.48550/arxiv.2402.06071
Apple’s machine learning researchers have developed an application that can accept a simple drawing and a text description to animate the drawing in the desired way. Tiffany Tseng, Ruijia Cheng and Jeffrey Nichols published an article describing the new application, called Keyframer, on the site arXiv preprint server.
AI researchers continue to find LLMs capable of accomplishing tasks that may not seem applicable at first glance. In this new effort, Apple researchers discovered that the GPT-4 LLM is capable of accepting both a simple drawing (in SVG format) and a text prompt describing a desired animation for the drawing, then producing an animation that executes the drawing. instructions.
For example, the application can accept an image of a rocket placed on a launch pad. The accompanying text may read: “Launch the rocket into the sky with a beam of fire blowing below.” The app will then attempt to animate the rocket to match the prompt.
The researchers explain that LLM can be used to animate the drawing by first imagining what needs to be done and then generating CSS animation code. As a result, the animation can be ported to other devices and the code can be modified to manually add or remove additional animation. New prompts can also be added.
They note that Keyframe also supports iterative animation, in which users can continue to add prompts to improve the finished product after each animation is generated, a way to produce the desired animated project in a short period of time.
Six images taken from animations generated by EP4 (HCLA) and EP9 (HCHA). The Saturn animation from Activity 1 of EP4 makes sparkles appear and disappear independently, clouds fade in, halos fade one after the other, and spots alternate colors from yellow to white. from orange to pink. The animation for Activity 2 Rocketship from EP9 moves the rocket up and down, the clouds grow and shrink, and the background elements (the moon and dots) move down to give the appearance of the rocket taking off. Credit: arXiv (2024). DOI: 10.48550/arxiv.2402.06071
In their paper, the researchers suggest that Keyframer could set the stage for transforming the animation landscape: if Apple made it available on all of its hardware platforms, for example, users would no longer need other applications. animation, and some could use it to create professional creations. products such as advertisements. It also allows non-professionals to create professional-quality products with limited effort.
More information:
Tiffany Tseng et al, Keyframer: Empowering Animation Design Using Large Language Models, arXiv (2024). DOI: 10.48550/arxiv.2402.06071
arXiv
© 2024 Science X Network
Quote: Apple’s Keyframer can animate simple drawings using text descriptions (February 16, 2024) retrieved February 16, 2024 from
This document is subject to copyright. Apart from fair use for private study or research purposes, no part may be reproduced without written permission. The content is provided for information only.
