29 Feb 2020 — Demos * Replicate Toggle. * Spaces Toggle. * Spaces Toggle.

: It can take the facial expressions or body movements from one person in a video and apply them to a completely different static image, such as making the Mona Lisa talk or dance.

[2003.00196] First Order Motion Model for Image Animation - arXiv

: The model doesn't need manual labels or prior knowledge of the object; it learns to identify keypoints and local transformations on its own.

The video filename video_2021-08-29_10-58-08.mp4 likely refers to a demonstration of an AI paper titled

This research was published by and colleagues at NeurIPS in 2019 and became highly viral in 2020 and 2021 for its ability to animate static images (like photos or paintings) using motion from a "driving" video. Key Features of the Paper

: It includes a special generator that can "fill in" parts of the background or object that become visible as the person moves, which was a major improvement over previous methods. Resources to Explore

Video_2021-08-29_10-58-08.mp4 Apr 2026

29 Feb 2020 — Demos * Replicate Toggle. * Spaces Toggle. * Spaces Toggle.

: It can take the facial expressions or body movements from one person in a video and apply them to a completely different static image, such as making the Mona Lisa talk or dance. video_2021-08-29_10-58-08.mp4

[2003.00196] First Order Motion Model for Image Animation - arXiv 29 Feb 2020 — Demos * Replicate Toggle

: The model doesn't need manual labels or prior knowledge of the object; it learns to identify keypoints and local transformations on its own. : It can take the facial expressions or

The video filename video_2021-08-29_10-58-08.mp4 likely refers to a demonstration of an AI paper titled

This research was published by and colleagues at NeurIPS in 2019 and became highly viral in 2020 and 2021 for its ability to animate static images (like photos or paintings) using motion from a "driving" video. Key Features of the Paper

: It includes a special generator that can "fill in" parts of the background or object that become visible as the person moves, which was a major improvement over previous methods. Resources to Explore