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

TL;DR

Extensive evaluations show that the 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.

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/.

Figures (10)

• Figure 1: We propose BiMotion, a fast, feed-forward B-spline–based method for dynamic 3D character generation. It produces continuous, high-quality expressive motion trajectories aligned with rich textual prompts, outperforming discrete temporal sampling-based methods such as AnimateAnyMesh wu2025animateanymesh under the same fixed-input constraint. See our project page for full motion dynamics.
• Figure 2: Overview. BiMotion uses a B-spline representation for motion generation. During training (red arrow), vertex differences are converted into control points and encoded into motion latents. During inference (black arrow), the initial mesh and the text prompt generate motion latents that are decoded into control points and converted into the generated mesh sequence via B-spline reprojection.
• Figure 3: B-spline VAE Pipeline. Given the initial shape $(\mathcal{P}_{n,0}, \mathcal{N}_0)$ and control points $\mathcal{P}_{n,k}$, the Encoder compresses them into latent codes $\mathbf{z}_0$ and $\mathbf{z}_{k}$. The Decoder reconstructs the predicted control points $\hat{\mathcal{P}}_{n,k}$, which are then reprojected to point differences via the B-spline basis. Note, * indicates that $\mathbf{F}_{k}'$ uses the FPS-sampled indices of $\mathbf{F}_0'$, $\oplus$ denotes matrix addition, and $\otimes$ denotes matrix multiplication.
• Figure 4: Qualitative Comparisons. Our method (BiMotion) results in superior motion quality and is more aligned with the user-provided text prompts. Artifacts for the baseline methods are highlighted in red. Please see the supplementary material for additional results.
• Figure 5: B-spline Ablation. B-spline interpolation from predicted control points achieves lower L1 error over the entire sequence than linear interpolation on sampled raw differences.