BiMotion: B-spline Motion for Text-guided Dynamic 3D Character Generation

2026-02-21

AI Summary

Existing text-driven dynamic 3D character animation methods are limited by fixed-length inputs and discrete frame representations, hindering the generation of semantically rich, coherent, and high-quality motions. This work proposes BiMotion, which models motion using continuous, differentiable B-spline curves and employs a closed-form Laplacian regularization to compress variable-length sequences into a fixed set of control points. To enhance motion detail and textual alignment, BiMotion introduces a normal fusion strategy, a correspondence-aware loss, and local rigidity constraints. Evaluated on the newly curated BIMO dataset, the method significantly improves motion expressiveness, generation quality, and text-motion consistency without altering the underlying generative architecture, while enabling efficient feedforward inference.

Technology Category

• Computer Vision: Motion & Tracking
• Intelligent Robots: Motion and Path Planning
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Text-guided dynamic 3D character generation has advanced rapidly, yet producing high-quality motion that faithfully reflects rich textual descriptions remains challenging. Existing methods tend to generate limited sub-actions or incoherent motion due to fixed-length temporal inputs and discrete frame-wise representations that fail to capture rich motion semantics. We address these limitations by representing motion with continuous differentiable B-spline curves, enabling more effective motion generation without modifying the capabilities of the underlying generative model. Specifically, our closed-form, Laplacian-regularized B-spline solver efficiently compresses variable-length motion sequences into compact representations with a fixed number of control points. Further, we introduce a normal-fusion strategy for input shape adherence along with correspondence-aware and local-rigidity losses for motion-restoration quality. To train our model, we collate BIMO, a new dataset containing diverse variable-length 3D motion sequences with rich, high-quality text annotations. Extensive evaluations show that our feed-forward framework BiMotion generates more expressive, higher-quality, and better prompt-aligned motions than existing state-of-the-art methods, while also achieving faster generation. Our project page is at: https://wangmiaowei.github.io/BiMotion.github.io/.

Problem

• text-guided motion generation
• dynamic 3D character
• motion semantics
• variable-length motion
• 3D animation

Innovation

• B-spline motion representation
• text-guided 3D character generation
• continuous motion modeling
• Laplacian-regularized spline solver
• variable-length motion compression