Animating Still Images
Kushagr Batra, Mridul Kavidayal

TL;DR
This paper introduces a deep learning-based method to animate still 2D images by segmenting, in-painting, and embedding the subject into a mesh for motion synthesis.
Contribution
It presents a novel pipeline combining segmentation, in-painting, and mesh embedding to animate static images, which was not previously achieved in this integrated manner.
Findings
Successfully animates still images with realistic motion.
Preserves background and image details during animation.
Demonstrates versatility across different image types.
Abstract
We present a method for imparting motion to a still 2D image. Our method uses deep learning to segment a section of the image denoted as subject, then uses in-painting to complete the background, and finally adds animation to the subject by embedding the image in a triangle mesh, while preserving the rest of the image.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsHuman Motion and Animation · Computer Graphics and Visualization Techniques · Advanced Vision and Imaging
