Apple has introduced Keyframer, a new artificial intelligence-driven animation generator. This generative tool utilizes a large language model capable of animating fixed vector images in response to natural language commands.
The Need for Improved Animation Tools
Existing AI image generators like Dall-E and Midjourney leverage textual descriptions; however, the creators noted that animation requires more complex parameters such as scene duration and object movement coordination. It is not easy to specify these thoroughly in one task description, necessitating alternative solutions like command refinement.
How Keyframer Works
This challenge is successfully addressed by Keyframer, which creates animated illustrations from static two-dimensional images. This AI uses the GPT-4 neural network to execute commands; SVG (Scalable Vector Graphic) images are used as an input, which the AI transforms into animated images using CSS code. Users can customize this code by manually editing it or providing additional textual commands.
Positive Reception
An initial showcase to animation designers and engineers yielded positive feedback. They recognized the potential of this tool to save time, noting that achieving similar results manually might require several hours.
Apple Continues to Invest in AI
This launch is another addition to a series of Apple’s AI-related research projects, like the recently released MGIE, a free image editor. Despite not positioning itself as an AI powerhouse, Apple nevertheless highlights its commitment to this field. CEO Tim Cook has promised a significant announcement on this front this year.
