用神经微分方程建模动态场景，实现长时间外推且内存不变。
Node-RF: Learning Generalized Continuous Space-Time Scene Dynamics with Neural ODE-based NeRFs
- 基于神经微分方程与动态NeRF结合，实现连续时空表征。
- 可在不增加内存前提下，对未观测轨迹进行长期外推。
- 适合需要泛化预测的动态场景建模任务，如自动驾驶模拟。
从视觉观测中预测场景动态极具挑战性。现有方法仅能捕捉观测边界内的动态，难以超出训练序列进行外推。Node-RF（基于神经微分方程的NeRF）通过将神经微分方程（NODEs）与动态神经辐射场（NeRFs）结合，实现了在恒定内存开销下的连续时间-空间表征，可推广至未观测轨迹。从视觉输入中，Node-RF学习隐式场景状态，通过求解微分方程随时间演化，并利用差分计算传播特征嵌入。基于NeRF的渲染器解析计算出的嵌入，合成任意视角以实现长距离外推。在多个共享动态的运动序列上联合训练，使模型具备对未见条件的泛化能力。实验表明，Node-RF无需显式模型即可刻画抽象系统行为，并识别未来预测的关键点。
原文摘要 · Abstract (English)
Predicting scene dynamics from visual observations is challenging. Existing methods capture dynamics only within observed boundaries failing to extrapolate far beyond the training sequence. Node-RF (Neural ODE-based NeRF) overcomes this limitation by integrating Neural Ordinary Differential Equations (NODEs) with dynamic Neural Radiance Fields (NeRFs), enabling a continuous-time, spatiotemporal representation that generalizes beyond observed trajectories at constant memory cost. From visual input, Node-RF learns an implicit scene state that evolves over time via an ODE solver, propagating feature embeddings via differential calculus. A NeRF-based renderer interprets calculated embeddings to synthesize arbitrary views for long-range extrapolation. Training on multiple motion sequences with shared dynamics allows for generalization to unseen conditions. Our experiments demonstrate that Node-RF can characterize abstract system behavior without explicit model to identify critical points for future predictions.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。