Newest papers

RayCo4D: Ray-Conditioned Continuous Motion Model for Monocular 4D Dynamic Reconstruction

Yuanxiang Ni, Xianliang Huang, Ruihan Hou, Feng Xie, Hao Zhang · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28351 · cited by 0 works (OpenAlex) · open-access PDF

Korean

동적 비디오에서 동적 3D 장면을 재구성하는 것은 소음과 간격으로 볼 수있는 관찰에서 움직임을 추론하는 것을 필요로합니다. 장거리 트랙은 가치있는 기하학적 제약을 제공하지만 공유, 지속적인 움직임으로 그들의 분열 된 시간 커버를 통합하는 것은 여전히 도전적입니다. 우리는 광선 조건화 된 monocular 4D 재구성을위한 움직임 모델인 RayCo4D를 소개합니다. 각 트랙은 카메라 라인에 제한된 참조 시간 3D 쿼리와 연관되어 있으며 이미지와 깊이 관찰은 현지 여러 조합을 통해 공유 늦은 움직임 노드를 제한합니다. 각 노드는 $\mathrm{SE}($3)의 시각적 측정을 명시적으로 분

Machine translation (Meta M2M100); the English original follows.

Abstract

Reconstructing dynamic 3D scenes from monocular video requires inferring motion from noisy and intermittently visible observations. Long-range tracks provide valuable geometric constraints, but integrating their fragmented temporal coverage into shared, continuous motion remains challenging. We introduce RayCo4D, a ray-conditioned continuous motion model for monocular 4D reconstruction that explicitly separates visual measurements from latent 3D motion states. Each track is associated with a reference-time 3D query constrained to its camera ray, and its image and depth observations constrain shared latent motion nodes through local many-to-many associations. Each node maintains an $\mathrm{SE}(3)$ pose and body-frame velocity at sparse support times. A white-noise-on-acceleration process prior couples neighboring states and defines continuous-time interpolation, allowing observations across visibility intervals to jointly constrain the same latent state sequence. The inferred shared motion is transferred to Gaussian primitives through dual-quaternion blending, while reference-anchored local $\mathrm{SE}(3)$ residuals capture fine-scale non-rigid deformation beyond the shared scaffold. Experiments on diverse monocular dynamic scenes show that RayCo4D achieves competitive novel-view synthesis quality, improves structural and temporal consistency under rapid motion and incomplete tracks, and supports continuous-time rendering.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer vision, Artificial intelligence, Computer science, Monocular, Motion (physics), Visibility

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange