AI painting tool Midjourney launches "pan" function, which can pan and expand scenes outside the picture

IT House reported on July 5 that the AI ​​painting tool Midjourney recently launched the "pan" function, which can generate scenes outside the picture. Users can control the "pan" of the picture by clicking "up, down, left, and right" in the input box.

▲ Picture source Midjourney’s Discord channel

It is reported that this function is mainly a refined version of the previous "extending the edges of the picture to bring about scene extension". Users can enter "/settings" in Midjourney's Discord channel to access the Remix mode to adjust various aspects outside the picture. A variable parameter that controls the generated screen content.

▲ Image expansion demonstration, source: Midjourney’s Discord channel

▲ Image expansion demonstration, source: Midjourney’s Discord channel

In Remix mode, users can upload their own photos, and the platform will crop the photos to a 1:1 square size. Then they can enter prompt words to create the content they need.

IT House found that although the user can always "pan" the image in a certain direction, once the first pan is passed, the prompt word cannot be modified to control the scene that will be generated next. In addition, users can only pan the image horizontally or vertically, not both, nor can they precisely adjust the specific panning amount, so developers need to make corresponding updates.

The above is the detailed content of AI painting tool Midjourney launches "pan" function, which can pan and expand scenes outside the picture. For more information, please follow other related articles on the PHP Chinese website!