MrNeRF (@janusch_patas)
2025-12-04 | โค๏ธ 247 | ๐ 33
Motion4D: Learning 3D-Consistent Motion and Semantics for 4D Scene Understanding
Contributions: โข We propose Motion4D, a model that integrates 2D priors from foundation models into a dynamic 3D Gaussian Splatting representation. This achieves consistent motion and semantic modeling from monocular videos.
โข We design a two-part iterative optimization framework comprising:
- Sequential optimization, which updates motion and semantic fields in consecutive stages to maintain local consistency.
- Global optimization, which jointly refines all attributes to ensure long-term coherence.
โข We introduce iterative motion refinement using 3D confidence maps and adaptive resampling to enhance dynamic scene reconstruction, alongside semantic refinement to correct 2D semantic inconsistencies through iterative updates with SAM2.
โข Our Motion4D significantly outperforms both 2D foundation models and existing 3D methods in tasks including video object segmentation, point-based tracking, and novel view synthesis.
๐ ์๋ณธ ๋งํฌ
๋ฏธ๋์ด
