Dynamic Novel View Synthesis in High Dynamic Range
Kaixuan Zhang, Zhipeng Xiong, Minxian Li, Mingwu Ren, Jiankang Deng, Xiatian Zhu
TL;DR
This work extends novel view synthesis to High Dynamic Range in dynamic scenes, formulating the HDR DNVS problem and introducing HDR-4DGS, a Gaussian Splatting–based framework with a dynamic tone mapper to bridge HDR and LDR domains over time. It incorporates a 4D radiance representation and a radiance bank with a Dynamic Radiance Context Learner to adapt tone mapping to evolving radiance distributions, enabling temporally coherent HDR renderings from arbitrary viewpoints and times. The authors also provide HDR-4D-Syn and HDR-4D-Real benchmarks and demonstrate state-of-the-art performance in HDR fidelity, temporal coherence, and efficiency on both synthetic and real data. Collectively, this work closes a critical gap between dynamic HDR reconstruction and practical rendering, with broad implications for immersive VR/AR content and photorealistic view synthesis under challenging lighting conditions.
Abstract
High Dynamic Range Novel View Synthesis (HDR NVS) seeks to learn an HDR 3D model from Low Dynamic Range (LDR) training images captured under conventional imaging conditions. Current methods primarily focus on static scenes, implicitly assuming all scene elements remain stationary and non-living. However, real-world scenarios frequently feature dynamic elements, such as moving objects, varying lighting conditions, and other temporal events, thereby presenting a significantly more challenging scenario. To address this gap, we propose a more realistic problem named HDR Dynamic Novel View Synthesis (HDR DNVS), where the additional dimension ``Dynamic'' emphasizes the necessity of jointly modeling temporal radiance variations alongside sophisticated 3D translation between LDR and HDR. To tackle this complex, intertwined challenge, we introduce HDR-4DGS, a Gaussian Splatting-based architecture featured with an innovative dynamic tone-mapping module that explicitly connects HDR and LDR domains, maintaining temporal radiance coherence by dynamically adapting tone-mapping functions according to the evolving radiance distributions across the temporal dimension. As a result, HDR-4DGS achieves both temporal radiance consistency and spatially accurate color translation, enabling photorealistic HDR renderings from arbitrary viewpoints and time instances. Extensive experiments demonstrate that HDR-4DGS surpasses existing state-of-the-art methods in both quantitative performance and visual fidelity. Source code will be released.