Learning Long-term Motion Embeddings for Efficient Kinematics Generation

TL;DR

This paper introduces a method for efficient long-term motion generation by learning a compressed motion embedding from large-scale trajectories, enabling realistic motion synthesis conditioned on text or spatial prompts.

Contribution

It proposes a novel long-term motion embedding learned via temporal compression, combined with a conditional flow-matching model for improved motion generation.

Findings

Motion embedding achieves 64x temporal compression.

Generated motions outperform state-of-the-art video models.

Method enables realistic long-term motion synthesis conditioned on prompts.

Abstract

Understanding and predicting motion is a fundamental component of visual intelligence. Although modern video models exhibit strong comprehension of scene dynamics, exploring multiple possible futures through full video synthesis remains prohibitively inefficient. We model scene dynamics orders of magnitude more efficiently by directly operating on a long-term motion embedding that is learned from large-scale trajectories obtained from tracker models. This enables efficient generation of long, realistic motions that fulfill goals specified via text prompts or spatial pokes. To achieve this, we first learn a highly compressed motion embedding with a temporal compression factor of 64x. In this space, we train a conditional flow-matching model to generate motion latents conditioned on task descriptions. The resulting motion distributions outperform those of both state-of-the-art video models and specialized task-specific approaches.