Learn how Microsoft’s new AI-powered tool, Pix2Gif, converts images to GIFs

Microsoft has been enhancing its AI capabilities for Copilot and other Windows apps. The research division has introduced a new AI model called Pix2Gif, which can transform still images into GIFs using an image translation technique. This allows users to quickly create funny and creative GIFs from their photos. The model also includes textual descriptions to provide prompts for the images.

What is Pix2Gif

Pix2Gif is a motion-controlled diffusion model that can convert images into GIFs. The tool not only captures information from the image, but also brings movement using text prompts and Motion Magnitude. Microsoft’s blog post emphasized that before training the AI model, “we carefully curate the data by extracting consistent image frames from the TGIF video text dataset, which provides rich information on subject changes over time. After pre-training, we apply our model with a null image to several video datasets.”

How to use Pix2Gif

Not that Pix2Gif is currently a research project and has not yet been made publicly available. However, Tom’s Guide reports that Pix2Gif can be used by downloading the tool locally for experimental purposes. Then you can upload a still image and text prompts for the desired animation to the tool. The tool then takes a few seconds to create a 2-second video or GIF based on the image processing and given description. Pix2Gif is still being tested and experimented with before the tool is made available to the public. However, the tool can offer several advantages when it comes to GIF and image manipulation.

Microsoft also said that Pix2Gif was trained using 100,000 animated gifs and subtitles. However, people are very excited to take advantage of AI video creation tools like OpenAI’ Sora, which will be released later this year.