[论文解读] Grow with the Flow: 4D Reconstruction of Growing Plants with Gaussian Flow Fields

Weihan Luo, Lily Goli+4 more|arXiv (Cornell University)|Feb 9, 2026

Plant Molecular Biology Research被引用 0

一句话总结

GrowFlow 将植物生长建模为一个由神经ODE控制的连续三维高斯流，能够从多视角时间间隔数据中实现生长植物的连贯、随时间延展的重建。它引入了带有3D高斯分布的逆向生长训练与时空 HexPlane 编码器来预测生长动力学。

ABSTRACT

Modeling the time-varying 3D appearance of plants during growth poses unique challenges: unlike most dynamic scenes, plants continuously generate new geometry as they expand, branch, and differentiate. Existing dynamic scene representations are ill-suited to this setting: deformation fields provide insufficient constraints to yield physically plausible scene dynamics, and 4D Gaussian splatting represents the same physical structures with different Gaussian primitives at different times, breaking temporal consistency. We introduce GrowFlow, a dynamic representation that couples 3D Gaussian primitives with a neural ordinary differential equation to model plant growth as a continuous flow field over geometric parameters (position, scale, and orientation). Our representation enables consistent appearance rendering and models nonlinear, continuous-time growth dynamics with full temporal correspondences for every primitive. To initialize a sufficient set of Gaussian primitives, we first reconstruct the mature plant and then learn a reverse-growth process, effectively simulating the plant's developmental history in reverse. GrowFlow achieves superior image quality and geometric coherence compared to prior methods on a new, multi-view timelapse dataset of plant growth, and provides the first temporally coherent representation for appearance modeling of growing 3D structures.

研究动机与目标

• 推动对随时间出现新几何形状的生长植物的高精度4D重建。
• 提出基于3D高斯基元的连续时间动态系统用于植物生长。
• 通过逆向生长和学得的流场实现随时间出现的新结构。
• 在保持单调生长假设的同时实现高光度和高几何精度。
• 在合成与真实多视角时间序列植物数据上展示最先进表现。

提出的方法

• 以随时间演化的几何参数的3D高斯斑点来表示场景。
• 使用控制高斯中心、旋转和尺度的速度场F_phi来建模时间变化。
• 结合时空 HexPlane 编码器和MLP解码器，预测每个高斯斑点的速度。
• 通过多阶段训练流程，包括边界重建、分段向后积分和全局优化。
• 采用逆向生长：学习在时间向后收缩高斯斑点，然后反向流以获得向前生长。

Figure 2 : Method overview. a) Our method first optimizes a set of 3D Gaussians on the fully-grown plant. b) Using the optimized 3D Gaussians from the fully-grown plant, we progressively train the dynamics model to learn the state of the plant at each timestep. After each reconstructed timestep, we

实验结果

研究问题

• RQ1 一个连续时间高斯流模型是否能够捕捉植物在时间上的非刚性、拓扑结构变化的生长？
• RQ2 将生长反向以收缩高斯斑点是否有利于可微分优化和稳定训练？
• RQ3 相较于最先进的基线，GrowFlow在合成与真实植物时序数据上的几何保真度和光度质量表现如何？
• RQ4 时空 HexPlane 编码是否对捕捉随时间的生长轨迹至关重要？

主要发现

方法	训练 PSNR (dB) ↑	训练 SSIM ↑	训练 LPIPS ↓	训练 CD ↓	插值 PSNR (dB) ↑	插值 SSIM ↑	插值 LPIPS ↓	插值 CD ↓	综合 PSNR (dB) ↑	综合 SSIM ↑	综合 LPIPS ↓	综合 CD ↓
4D-GS	33.04	0.946	0.094	0.73	32.77	0.944	0.094	0.78	32.81	0.944	0.094	0.77
4DGS	30.19	0.939	0.107	12.00	29.11	0.905	0.145	11.95	29.29	0.910	0.138	11.96
Dynamic 3DGS	32.48	0.912	0.154	13.18	32.03	0.908	0.158	13.64	32.11	0.909	0.157	13.56
Ours	35.43	0.957	0.065	0.10	34.93	0.955	0.066	0.11	35.02	0.956	0.066	0.11

• GrowFlow 在 seen 与插值时间步上实现了几何上连贯的生长轨迹，具有高时间一致性。
• 在合成数据上，GrowFlow在几何精度（Chamfer Distance）和图像质量（PSNR、SSIM、LPIPS）方面优于基线。
• 在真实场景捕获数据上，GrowFlow 的新视图渲染更优，插值稳定性高，基线难以做到。
• 基于 HexPlane 的时空编码器与边界重建阶段对性能至关重要；替换它们会降低结果。
• 逆向生长训练使得新几何形状的连续引入成为可能，无需不可微的附加项。

Figure 3 : Results on synthetic data. We compare results on both seen and interpolated times. GrowFlow achieves stable, coherent geometry, unlike prior methods that show visually correct RGB but inconsistent deformations. Yellow marks interpolated frames.