Multi-Object Sketch Animation with Grouping and Motion Trajectory Priors

2025-08-21

AI Summary

Existing vector sketch animation methods struggle to model multi-object interactions and complex motion, suffering from limitations such as single-object constraints, temporal inconsistency, and poor generalization. To address these issues, we propose GroupSketch: a novel framework introducing semantic grouping and a Group-based Displacement Network (GDN) that decomposes multi-object sketches into semantically coherent groups and jointly models their motion. GroupSketch adopts a two-stage pipeline—Motion Initialization (semantic grouping + keyframe interpolation) and Motion Refinement (incorporating motion priors from text-to-video diffusion models and a Contextual Context Fusion Enhancement module, CCFE)—to significantly improve temporal coherence and dynamic plausibility. Extensive experiments demonstrate that GroupSketch substantially outperforms state-of-the-art methods on multi-object sketch animation, achieving superior generation quality and strong generalization across diverse sketch styles and object configurations, thereby expanding the practical applicability of sketch-driven animation.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Generation
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

We introduce GroupSketch, a novel method for vector sketch animation that effectively handles multi-object interactions and complex motions. Existing approaches struggle with these scenarios, either being limited to single-object cases or suffering from temporal inconsistency and poor generalization. To address these limitations, our method adopts a two-stage pipeline comprising Motion Initialization and Motion Refinement. In the first stage, the input sketch is interactively divided into semantic groups and key frames are defined, enabling the generation of a coarse animation via interpolation. In the second stage, we propose a Group-based Displacement Network (GDN), which refines the coarse animation by predicting group-specific displacement fields, leveraging priors from a text-to-video model. GDN further incorporates specialized modules, such as Context-conditioned Feature Enhancement (CCFE), to improve temporal consistency. Extensive experiments demonstrate that our approach significantly outperforms existing methods in generating high-quality, temporally consistent animations for complex, multi-object sketches, thus expanding the practical applications of sketch animation.

Problem

• Handles multi-object interactions in sketch animation
• Addresses temporal inconsistency in complex motions
• Improves generalization for multi-object sketch scenarios

Innovation

• Two-stage pipeline with motion initialization
• Group-based Displacement Network for refinement
• Context-conditioned Feature Enhancement module