E-4DGS: High-Fidelity Dynamic Reconstruction from the Multi-view Event Cameras

2025/08/13 by Chaoran Feng, Feng, Chaoran, Zhenyu Tang +13 · 3 citations
Computer Science · Engineering · Physics and Astronomy · #Advanced Data Storage Technologies #Advanced Memory and Neural Computing #Advanced Optical Sensing Technologies #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2508.09912

openalex publication_date 2025/08/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Novel view synthesis and 4D reconstruction techniques predominantly rely on RGB cameras, thereby inheriting inherent limitations such as the dependence on adequate lighting, susceptibility to motion blur, and a limited dynamic range. Event cameras, offering advantages of low power, high temporal resolution and high dynamic range, have brought a new perspective to addressing the scene reconstruction challenges in high-speed motion and

Citations

• NeuralGS: Bridging Neural Fields and 3D Gaussian Splatting for Compact 3D Representations
• A Survey on Event-driven 3D Reconstruction: Development under Different Categories
• Swift4D:Adaptive divide-and-conquer Gaussian Splatting for compact and efficient reconstruction of dynamic scene
• Deformable Gaussian Splatting for Efficient and High-Fidelity Reconstruction of Surgical Scenes
• 4D Gaussian Splatting: Modeling Dynamic Scenes with Native 4D Primitives
• Next Patch Prediction for Autoregressive Visual Generation
• SweepEvGS: Event-Based 3D Gaussian Splatting for Macro and Micro Radiance Field Rendering from a Single Sweep
• Deblur4DGS: 4D Gaussian Splatting from Blurry Monocular Video
• Template-free Articulated Gaussian Splatting for Real-time Reposable Dynamic View Synthesis
• DreamDance: Animating Human Images by Enriching 3D Geometry Cues from 2D Poses
• USP-Gaussian: Unifying Spike-based Image Reconstruction, Pose Correction and Gaussian Splatting
• 4D Gaussian Splatting with Scale-aware Residual Field and Adaptive Optimization for Real-time Rendering of Temporally Complex Dynamic Scenes
• E-3DGS: Gaussian Splatting with Exposure and Motion Events
• DN-4DGS: Denoised Deformable Network with Temporal-Spatial Aggregation for Dynamic Scene Rendering
• IncEventGS: Pose-Free Gaussian Splatting from a Single Event Camera
• MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion
• Elite-EvGS: Learning Event-based 3D Gaussian Splatting by Distilling Event-to-Video Priors
• CoSEC: A Coaxial Stereo Event Camera Dataset for Autonomous Driving
• Cycle3D: High-quality and Consistent Image-to-3D Generation via Generation-Reconstruction Cycle
• Shape of Motion: 4D Reconstruction from a Single Video
• Ev-GS: Event-based Gaussian splatting for Efficient and Accurate Radiance Field Rendering
• EvaGaussians: Event Stream Assisted Gaussian Splatting from Blurry Images
• EvGGS: A Collaborative Learning Framework for Event-based Generalizable Gaussian Splatting
• MagicTime: Time-lapse Video Generation Models as Metamorphic Simulators
• 4D-Rotor Gaussian Splatting: Towards Efficient Novel View Synthesis for Dynamic Scenes
• Repaint123: Fast and High-quality One Image to 3D Generation with Progressive Controllable 2D Repainting
• Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle
• FreestyleRet: Retrieving Images from Style-Diversified Queries
• LEOD: Label-Efficient Object Detection for Event Cameras
• LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPS
• Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction
• Robust e-NeRF: NeRF from Sparse & Noisy Events under Non-Uniform Motion
• 3D Gaussian Splatting for Real-Time Radiance Field Rendering
• 3D Gaussian Splatting for Real-Time Radiance Field Rendering
• NeRF-DS: Neural Radiance Fields for Dynamic Specular Objects
• ULIP: Learning a Unified Representation of Language, Images, and Point Clouds for 3D Understanding
• E-NeRF: Neural Radiance Fields from a Moving Event Camera
• TimeReplayer: Unlocking the Potential of Event Cameras for Video Interpolation
• TUM-VIE: The TUM Stereo Visual-Inertial Event Dataset
• EventHPE: Event-based 3D Human Pose and Shape Estimation
• DSEC: A Stereo Event Camera Dataset for Driving Scenarios
• D-NeRF: Neural Radiance Fields for Dynamic Scenes
• Nerfies: Deformable Neural Radiance Fields
• Bringing a Blurry Frame Alive at High Frame-Rate with an Event Camera
• Event-boosted Deformable 3D Gaussians for Dynamic Scene Reconstruction

Cited by

• Breaking the Vicious Cycle: Coherent 3D Gaussian Splatting from Sparse and Motion-Blurred Views
• MultiMotion: Multi Subject Video Motion Transfer via Video Diffusion Transformer
• ContextFlow: Training-Free Video Object Editing via Adaptive Context Enrichment
• Dark-EvGS: Event Camera as an Eye for Radiance Field in the Dark

Related

• 4DSloMo: 4D Reconstruction for High Speed Scene with Asynchronous Capture
• E-3DGS: Event-Based Novel View Rendering of Large-Scale Scenes Using 3D Gaussian Splatting
• Dynamic EventNeRF: Reconstructing General Dynamic Scenes from Multi-view RGB and Event Streams
• Any4D: Unified Feed-Forward Metric 4D Reconstruction
• Event Camera Guided Visual Media Restoration & 3D Reconstruction: A Survey