分而治之：重新思考神经辐射场的训练范式
Divide and Conquer: Rethinking the Training Paradigm of Neural Radiance Fields
AI总结：
针对NeRF标准训练范式假设训练图像重要性均等导致复杂几何视图渲染性能不佳的问题，提出分而治之的训练框架，按视觉相似性分组训练专用模型再经师生蒸馏聚合，在两个公开数据集上提升了基线模型渲染质量并收敛到更优最小值。
摘要：
神经辐射场（NeRFs）在合成3D场景的高保真视图方面展现出潜力，但NeRF的标准训练范式预设训练集中每张图像的重要性相等。这一假设对渲染呈现复杂几何结构的特定视图构成了重大挑战，进而导致性能未达最优。本文深入研究了当前训练范式的影响，并对其进行重新设计，以提升NeRF的渲染质量。我们根据输入视图的视觉相似性将其划分为多个组，针对每组训练单独的模型，使每个模型能专注于特定区域，且不会牺牲速度或效率。随后，通过师生蒸馏范式将这些专用模型的知识聚合到单个模型中，实现在线渲染的空间高效性。在实验方面，我们在两个公开数据集（即NeRF synthetic和Tanks&Temples）上评估了这一新型训练框架。评估结果表明，我们的DaC训练流水线提升了当前最优基线模型的渲染质量，同时能收敛到更优的最小值。
英文摘要：
Neural radiance fields (NeRFs) have exhibited potential in synthesizing high-fidelity views of 3D scenes but the standard training paradigm of NeRF presupposes an equal importance for each image in the training set. This assumption poses a significant challenge for rendering specific views presenting intricate geometries, thereby resulting in suboptimal performance. In this paper, we take a closer look at the implications of the current training paradigm and redesign this for more superior rendering quality by NeRFs. Dividing input views into multiple groups based on their visual similarities and training individual models on each of these groups enables each model to specialize on specific regions without sacrificing speed or efficiency. Subsequently, the knowledge of these specialized models is aggregated into a single entity via a teacher-student distillation paradigm, enabling spatial efficiency for online render-ing. Empirically, we evaluate our novel training framework on two publicly available datasets, namely NeRF synthetic and Tanks&Temples. Our evaluation demonstrates that our DaC training pipeline enhances the rendering quality of a state-of-the-art baseline model while exhibiting convergence to a superior minimum.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
