Mono4DGS-HDR: High Dynamic Range 4D Gaussian Splatting from Alternating-exposure Monocular Videos

2025-10-21

AI Summary

This work addresses the problem of reconstructing 4D high-dynamic-range (HDR) scenes from unposed, monocular, alternating-exposure low-dynamic-range (LDR) video sequences. We propose the first renderable 4D HDR scene reconstruction method based on a Gaussian splatting representation, featuring a two-stage optimization framework: geometric and HDR radiance initialization in an orthographic camera coordinate space, followed by joint refinement in the world coordinate system. To ensure temporal coherence and stability in HDR rendering, we introduce a time-aware brightness regularization term. Our key contribution is the first monocular video-based 4D HDR Gaussian reconstruction without requiring pose priors—enabling robust novel-view synthesis for dynamic scenes. Evaluated on a newly established HDR dynamic benchmark, our method achieves significant improvements over state-of-the-art approaches in rendering accuracy, temporal consistency, and inference speed.

Technology Category

• Computer Vision: 3D Computer Vision
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We introduce Mono4DGS-HDR, the first system for reconstructing renderable 4D high dynamic range (HDR) scenes from unposed monocular low dynamic range (LDR) videos captured with alternating exposures. To tackle such a challenging problem, we present a unified framework with two-stage optimization approach based on Gaussian Splatting. The first stage learns a video HDR Gaussian representation in orthographic camera coordinate space, eliminating the need for camera poses and enabling robust initial HDR video reconstruction. The second stage transforms video Gaussians into world space and jointly refines the world Gaussians with camera poses. Furthermore, we propose a temporal luminance regularization strategy to enhance the temporal consistency of the HDR appearance. Since our task has not been studied before, we construct a new evaluation benchmark using publicly available datasets for HDR video reconstruction. Extensive experiments demonstrate that Mono4DGS-HDR significantly outperforms alternative solutions adapted from state-of-the-art methods in both rendering quality and speed.

Problem

• Reconstructing 4D HDR scenes from alternating-exposure monocular videos
• Eliminating camera pose requirements for robust HDR video reconstruction
• Enhancing temporal consistency of HDR appearance in 4D scenes

Innovation

• Alternating-exposure monocular videos for HDR reconstruction
• Two-stage Gaussian Splatting optimization without camera poses
• Temporal luminance regularization for consistent HDR appearance