Search results: clip

Meta uses AI for a new system that generates videos from text descriptions

Meta has presented "Make-A-Video" a new application that, based on Artificial Intelligence, is capable to generate a short video from a text description.  The AI ​​system used is based on imaging technology that learns from paired text and image data and videos without context. From them, the new system is able to create a high-quality video,  converting the provided text into a short clip.  According to the statement issued from Mark Zuckerberg's company, "Make-A-Video" represents an important advance with respect to "Make-A-Scene", the similar system presented by Meta last summer, that generates precise images from descriptions of text and sketches.