Director: Instance-aware Gaussian Splatting for Dynamic Scene Modeling and Understanding

• URL: http://arxiv.org/abs/2604.01678v1
• Date: Thu, 02 Apr 2026 06:29:53 GMT
• Title: Director: Instance-aware Gaussian Splatting for Dynamic Scene Modeling and Understanding
• Abstract summary: Volumetric seeks to model rendering scenes as temporally coherent 4D representations.<n>Director achieves temporally coherent 4D reconstructions while simultaneously enabling instance segmentation and open-vocabulary querying.
• Score: 40.60302826781783
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Volumetric video seeks to model dynamic scenes as temporally coherent 4D representations. While recent Gaussian-based approaches achieve impressive rendering fidelity, they primarily emphasize appearance but are largely agnostic to instance-level structure, limiting stable tracking and semantic reasoning in highly dynamic scenarios. In this paper, we present Director, a unified spatio-temporal Gaussian representation that jointly models human performance, high-fidelity rendering, and instance-level semantics. Our key insight is that embedding instance-consistent semantics naturally complements 4D modeling, enabling more accurate scene decomposition while supporting robust dynamic scene understanding. To this end, we leverage temporally aligned instance masks and sentence embeddings derived from Multimodal Large Language Models to supervise the learnable semantic features of each Gaussian via two MLP decoders, enabling language-aligned 4D representations and enforcing identity consistency over time. To enhance temporal stability, we bridge 2D optical flow with 4D Gaussians and finetune their motions, yielding reliable initialization and reducing drift. For the training, we further introduce a geometry-aware SDF constraints, along with regularization terms that enforces surface continuity, enhancing temporal coherence in dynamic foreground modeling. Experiments demonstrate that Director achieves temporally coherent 4D reconstructions while simultaneously enabling instance segmentation and open-vocabulary querying.

Related papers

• Multi4D: High-Fidelity Dynamic Gaussian Splatting via Multi-Level Competitive Allocation [11.135752651274663]
  We introduce Multi4D, a framework for high-fidelity dynamic Gaussian splatting.<n>We distribute modeling capacity across three levels: static structure, persistent dynamic geometry, and transient appearance primitives.<n>This allocation preserves long-term motion consistency while capturing fine dynamic detail, achieving state-of-the-art rendering quality and real-time performance.
  arXiv  Detail & Related papers  (2026-06-20T19:11:26Z)
• Thinking in Dynamics: How Multimodal Large Language Models Perceive, Track, and Reason Dynamics in Physical 4D World [49.80040477190479]
  Humans inhabit a physical 4D world where geometric structure and semantic content evolve over time.<n>We introduce Dyn-Bench, a large-scale benchmark built from diverse real-world and synthetic video datasets.<n>We find that existing models cannot simultaneously maintain strong performance in both-temporal reasoning and dynamic object grounding.
  arXiv  Detail & Related papers  (2026-03-13T07:42:16Z)
• Mango-GS: Enhancing Spatio-Temporal Consistency in Dynamic Scenes Reconstruction using Multi-Frame Node-Guided 4D Gaussian Splatting [17.957935001501106]
  We present Mango-GS, a node-guided framework for high-fidelity 4D reconstruction.<n>We leverage a temporal Transformer to model motion dependencies within a short window of frames.<n>Mango-GS achieves state-of-the-art reconstruction quality and real-time rendering speed.
  arXiv  Detail & Related papers  (2026-03-12T05:00:16Z)
• SS4D: Native 4D Generative Model via Structured Spacetime Latents [50.29500511908054]
  We present SS4D, a native 4D generative model that synthesizes dynamic 3D objects directly from monocular video.<n>We train a generator directly on 4D data, achieving high fidelity, temporal coherence, and structural consistency.
  arXiv  Detail & Related papers  (2025-12-16T10:45:06Z)
• MoRel: Long-Range Flicker-Free 4D Motion Modeling via Anchor Relay-based Bidirectional Blending with Hierarchical Densification [10.799902862870288]
  MoRel is a novel framework for temporally consistent and memory-efficient modeling of long-range dynamic scenes.<n>Our approach mitigates temporal discontinuities and flickering artifacts.<n>It achieves temporally coherent and flicker-free long-range 4D reconstruction while maintaining bounded memory usage.
  arXiv  Detail & Related papers  (2025-12-10T02:49:09Z)
• Motion4D: Learning 3D-Consistent Motion and Semantics for 4D Scene Understanding [54.859943475818234]
  We present Motion4D, a novel framework that integrates 2D priors from foundation models into a unified 4D Gaussian Splatting representation.<n>Our method features a two-part iterative optimization framework: 1) Sequential optimization, which updates motion and semantic fields in consecutive stages to maintain local consistency, and 2) Global optimization, which jointly refines all attributes for long-term coherence.<n>Our method significantly outperforms both 2D foundation models and existing 3D-based approaches across diverse scene understanding tasks, including point-based tracking, video object segmentation, and novel view synthesis.
  arXiv  Detail & Related papers  (2025-12-03T09:32:56Z)
• SCas4D: Structural Cascaded Optimization for Boosting Persistent 4D Novel View Synthesis [53.10680153186481]
  We propose SCas4D, a cascaded optimization framework that leverages structural patterns in 3D Gaussian Splatting for dynamic scenes.<n>By progressively refining deformations from coarse part-level to fine point-level, SCas4D achieves convergence within 100 iterations per time frame.<n>The approach also demonstrates effectiveness in self-supervised articulated object segmentation, novel view synthesis, and dense point tracking tasks.
  arXiv  Detail & Related papers  (2025-10-08T06:39:33Z)
• CoDa-4DGS: Dynamic Gaussian Splatting with Context and Deformation Awareness for Autonomous Driving [12.006435326659526]
  We introduce a novel 4D Gaussian Splatting (4DGS) approach to improve dynamic scene rendering.<n> Specifically, we employ a 2D semantic segmentation foundation model to self-supervise the 4D semantic features of Gaussians.<n>By aggregating and encoding both semantic and temporal deformation features, each Gaussian is equipped with cues for potential deformation compensation.
  arXiv  Detail & Related papers  (2025-03-09T19:58:51Z)
• 4D Gaussian Splatting: Modeling Dynamic Scenes with Native 4D Primitives [115.67081491747943]
  Dynamic 3D scene representation and novel view synthesis are crucial for enabling AR/VR and metaverse applications.<n>We reformulate the reconstruction of a time-varying 3D scene as approximating its underlying 4D volume.<n>We derive several compact variants that effectively reduce the memory footprint to address its storage bottleneck.
  arXiv  Detail & Related papers  (2024-12-30T05:30:26Z)
• UrbanGS: Semantic-Guided Gaussian Splatting for Urban Scene Reconstruction [86.4386398262018]
  UrbanGS uses 2D semantic maps and an existing dynamic Gaussian approach to distinguish static objects from the scene.<n>For potentially dynamic objects, we aggregate temporal information using learnable time embeddings.<n>Our approach outperforms state-of-the-art methods in reconstruction quality and efficiency.
  arXiv  Detail & Related papers  (2024-12-04T16:59:49Z)
• Motion2VecSets: 4D Latent Vector Set Diffusion for Non-rigid Shape Reconstruction and Tracking [52.393359791978035]
  Motion2VecSets is a 4D diffusion model for dynamic surface reconstruction from point cloud sequences. We parameterize 4D dynamics with latent sets instead of using global latent codes. For more temporally-coherent object tracking, we synchronously denoise deformation latent sets and exchange information across multiple frames.
  arXiv  Detail & Related papers  (2024-01-12T15:05:08Z)
• Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting [8.078460597825142]
  Reconstructing dynamic 3D scenes from 2D images and generating diverse views over time is challenging due to scene complexity and temporal dynamics. We propose to approximate the underlying-temporal rendering volume of a dynamic scene by optimizing a collection of 4D primitives, with explicit geometry and appearance modeling. Our model is conceptually simple, consisting of a 4D Gaussian parameterized by anisotropic ellipses that can rotate arbitrarily in space and time, as well as view-dependent and time-evolved appearance represented by the coefficient of 4D spherindrical harmonics.
  arXiv  Detail & Related papers  (2023-10-16T17:57:43Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.