最近，神经辐射场（NERF）正在彻底改变新型视图合成（NVS）的卓越性能。但是，NERF及其变体通常需要进行冗长的每场训练程序，其中将多层感知器（MLP）拟合到捕获的图像中。为了解决挑战，已经提出了体素网格表示，以显着加快训练的速度。但是，这些现有方法只能处理静态场景。如何开发有效，准确的动态视图合成方法仍然是一个开放的问题。将静态场景的方法扩展到动态场景并不简单，因为场景几何形状和外观随时间变化。在本文中，基于素素网格优化的最新进展，我们提出了一种快速变形的辐射场方法来处理动态场景。我们的方法由两个模块组成。第一个模块采用变形网格来存储3D动态功能，以及使用插值功能将观测空间中的3D点映射到规范空间的变形的轻巧MLP。第二个模块包含密度和颜色网格，以建模场景的几何形状和密度。明确对阻塞进行了建模，以进一步提高渲染质量。实验结果表明，我们的方法仅使用20分钟的训练就可以实现与D-NERF相当的性能，该训练比D-NERF快70倍以上，这清楚地证明了我们提出的方法的效率。

3D reconstruction and novel view synthesis of dynamic scenes from collections of single views recently gained increased attention. Existing work shows impressive results for synthetic setups and forward-facing real-world data, but is severely limited in the training speed and angular range for generating novel views. This paper addresses these limitations and proposes a new method for full 360{\deg} novel view synthesis of non-rigidly deforming scenes. At the core of our method are: 1) An efficient deformation module that decouples the processing of spatial and temporal information for acceleration at training and inference time; and 2) A static module representing the canonical scene as a fast hash-encoded neural radiance field. We evaluate the proposed approach on the established synthetic D-NeRF benchmark, that enables efficient reconstruction from a single monocular view per time-frame randomly sampled from a full hemisphere. We refer to this form of inputs as monocularized data. To prove its practicality for real-world scenarios, we recorded twelve challenging sequences with human actors by sampling single frames from a synchronized multi-view rig. In both cases, our method is trained significantly faster than previous methods (minutes instead of days) while achieving higher visual accuracy for generated novel views. Our source code and data is available at our project page https://graphics.tu-bs.de/publications/kappel2022fast.

神经辐射场（NERF）在建模3D场景和合成新型视图图像方面取得了巨大成功。但是，大多数以前的NERF方法需要大量时间来优化一个场景。显式数据结构，例如体素特征，显示出加速训练过程的巨大潜力。但是，体素特征面临两个大挑战，要应用于动态场景，即建模时间信息并捕获不同的点运动尺度。我们通过用时间感知的体素特征（称为Tineuvox）表示场景来提出一个辐射现场框架。引入了一个微小的坐标变形网络，以模拟粗糙运动轨迹，并在辐射网络中进一步增强了时间信息。提出了一种多距离插值方法，并应用于体素特征，以模拟小运动和大型运动。我们的框架大大加快了动态光芒度场的优化，同时保持高渲染质量。经验评估均在合成场景和真实场景上进行。我们的Tineuvox仅需8分钟和8 MB的存储成本即可完成培训，同时表现出比以前的动态NERF方法相似甚至更好的渲染性能。

Point of View & TimeFigure 1: We propose D-NeRF, a method for synthesizing novel views, at an arbitrary point in time, of dynamic scenes with complex non-rigid geometries. We optimize an underlying deformable volumetric function from a sparse set of input monocular views without the need of ground-truth geometry nor multi-view images. The figure shows two scenes under variable points of view and time instances synthesised by the proposed model.

Photo-realistic free-viewpoint rendering of real-world scenes using classical computer graphics techniques is challenging, because it requires the difficult step of capturing detailed appearance and geometry models. Recent studies have demonstrated promising results by learning scene representations that implicitly encode both geometry and appearance without 3D supervision. However, existing approaches in practice often show blurry renderings caused by the limited network capacity or the difficulty in finding accurate intersections of camera rays with the scene geometry. Synthesizing high-resolution imagery from these representations often requires time-consuming optical ray marching. In this work, we introduce Neural Sparse Voxel Fields (NSVF), a new neural scene representation for fast and high-quality free-viewpoint rendering. NSVF defines a set of voxel-bounded implicit fields organized in a sparse voxel octree to model local properties in each cell. We progressively learn the underlying voxel structures with a diffentiable ray-marching operation from only a set of posed RGB images. With the sparse voxel octree structure, rendering novel views can be accelerated by skipping the voxels containing no relevant scene content. Our method is typically over 10 times faster than the state-of-the-art (namely, NeRF (Mildenhall et al., 2020)) at inference time while achieving higher quality results. Furthermore, by utilizing an explicit sparse voxel representation, our method can easily be applied to scene editing and scene composition. We also demonstrate several challenging tasks, including multi-scene learning, free-viewpoint rendering of a moving human, and large-scale scene rendering. Code and data are available at our website: https://github.com/facebookresearch/NSVF.

本文旨在减少透明辐射场的渲染时间。一些最近的作品用图像编码器配备了神经辐射字段，能够跨越场景概括，这避免了每场景优化。但是，它们的渲染过程通常很慢。主要因素是，在推断辐射场时，它们在空间中的大量点。在本文中，我们介绍了一个混合场景表示，它结合了最佳的隐式辐射场和显式深度映射，以便有效渲染。具体地，我们首先构建级联成本量，以有效地预测场景的粗糙几何形状。粗糙几何允许我们在场景表面附近的几个点来样，并显着提高渲染速度。该过程是完全可疑的，使我们能够仅从RGB图像共同学习深度预测和辐射现场网络。实验表明，该方法在DTU，真正的前瞻性和NERF合成数据集上展示了最先进的性能，而不是比以前的最可推广的辐射现场方法快至少50倍。我们还展示了我们的方法实时综合动态人类执行者的自由观点视频。代码将在https://zju3dv.github.io/enerf/处提供。

我们提出了逐渐变化的辐射场（PDRF），这是一种从模糊图像中有效重建高质量辐射场的新方法。虽然当前的最先进的（SOTA）场景重建方法实现了光真实的渲染，因此清洁源视图会导致其性能在源视图受模糊影响的影响时会受到影响，这通常是野外图像的观察。以前的脱毛方法要么不考虑3D几何形状，要么是计算强度。为了解决这些问题，PDRF是Radiance Field建模中逐渐消除的方案，通过合并3D场景上下文来准确地模拟模糊。 PDRF进一步使用了有效的重要性采样方案，从而导致快速场景优化。具体而言，PDRF提出了一个粗射线渲染器，以快速估计体素密度和特征。然后，使用精细的体素渲染器来实现高质量的射线追踪。我们执行广泛的实验，并表明PDRF比以前的SOTA快15倍，同时在合成场景和真实场景上都取得更好的性能。

Figure 1. Given a monocular image sequence, NR-NeRF reconstructs a single canonical neural radiance field to represent geometry and appearance, and a per-time-step deformation field. We can render the scene into a novel spatio-temporal camera trajectory that significantly differs from the input trajectory. NR-NeRF also learns rigidity scores and correspondences without direct supervision on either. We can use the rigidity scores to remove the foreground, we can supersample along the time dimension, and we can exaggerate or dampen motion.

我们提出了高动态范围辐射（HDR）字段，HDR-PLENOXELS，它学习了3D HDR辐射场的肺化功能，几何信息和2D低动态范围（LDR）图像中固有的不同摄像机设置。我们基于体素的卷渲染管道可重建HDR辐射字段，仅以端到端的方式从不同的相机设置中拍摄的多视图LDR图像，并且具有快速的收敛速度。为了在现实世界中处理各种摄像机，我们引入了一个音调映射模块，该模块模拟了数字相机内成像管道（ISP）（ISP）和DISTANGLES辐射测定设置。我们的音调映射模块可以通过控制每个新型视图的辐射设置来渲染。最后，我们构建一个具有不同摄像机条件的多视图数据集，适合我们的问题设置。我们的实验表明，HDR-Plenoxels可以从具有各种相机的LDR图像中表达细节和高质量的HDR新型视图。

我们介绍了一种超快速的收敛方法来重建从一组图像中捕获具有已知姿势的场景的图像的每场辐射场。该任务通常适用于新颖的视图综合，最近是由神经辐射领域（NERF）彻底改革为其最先进的质量和灵活性。然而，NERF及其变体需要漫长的训练时间来为单个场景的数小时到几天。相比之下，我们的方法实现了NERF相当的质量，并通过单个GPU在不到15分钟内从划痕中迅速收敛。我们采用由密度体素网格组成的表示，用于场景几何形状和具有浅网络的特征体素网格，用于复杂的视图依赖性外观。用明确和离散化卷表示的建模并不是新的，但我们提出了两种简单而非琐碎的技术，有助于快速收敛速度和高质量的输出。首先，我们介绍了体素密度的激活后插值，其能够以较低的网格分辨率产生尖锐的表面。其次，直接体素密度优化容易发生次优几何解决方案，因此我们通过施加多个前沿来强制优化过程。最后，对五个内向的基准评估表明，我们的方法匹配，如果没有超越Nerf的质量，但它只需15分钟即可从头开始训练新场景。

Synthesizing high-fidelity videos from real-world multi-view input is challenging because of the complexities of real-world environments and highly dynamic motions. Previous works based on neural radiance fields have demonstrated high-quality reconstructions of dynamic scenes. However, training such models on real-world scenes is time-consuming, usually taking days or weeks. In this paper, we present a novel method named MixVoxels to better represent the dynamic scenes with fast training speed and competitive rendering qualities. The proposed MixVoxels represents the 4D dynamic scenes as a mixture of static and dynamic voxels and processes them with different networks. In this way, the computation of the required modalities for static voxels can be processed by a lightweight model, which essentially reduces the amount of computation, especially for many daily dynamic scenes dominated by the static background. To separate the two kinds of voxels, we propose a novel variation field to estimate the temporal variance of each voxel. For the dynamic voxels, we design an inner-product time query method to efficiently query multiple time steps, which is essential to recover the high-dynamic motions. As a result, with 15 minutes of training for dynamic scenes with inputs of 300-frame videos, MixVoxels achieves better PSNR than previous methods. Codes and trained models are available at https://github.com/fengres/mixvoxels

We present a method that achieves state-of-the-art results for synthesizing novel views of complex scenes by optimizing an underlying continuous volumetric scene function using a sparse set of input views. Our algorithm represents a scene using a fully-connected (nonconvolutional) deep network, whose input is a single continuous 5D coordinate (spatial location (x, y, z) and viewing direction (θ, φ)) and whose output is the volume density and view-dependent emitted radiance at that spatial location. We synthesize views by querying 5D coordinates along camera rays and use classic volume rendering techniques to project the output colors and densities into an image. Because volume rendering is naturally differentiable, the only input required to optimize our representation is a set of images with known camera poses. We describe how to effectively optimize neural radiance fields to render photorealistic novel views of scenes with complicated geometry and appearance, and demonstrate results that outperform prior work on neural rendering and view synthesis. View synthesis results are best viewed as videos, so we urge readers to view our supplementary video for convincing comparisons.

Recent methods for neural surface representation and rendering, for example NeuS, have demonstrated remarkably high-quality reconstruction of static scenes. However, the training of NeuS takes an extremely long time (8 hours), which makes it almost impossible to apply them to dynamic scenes with thousands of frames. We propose a fast neural surface reconstruction approach, called NeuS2, which achieves two orders of magnitude improvement in terms of acceleration without compromising reconstruction quality. To accelerate the training process, we integrate multi-resolution hash encodings into a neural surface representation and implement our whole algorithm in CUDA. We also present a lightweight calculation of second-order derivatives tailored to our networks (i.e., ReLU-based MLPs), which achieves a factor two speed up. To further stabilize training, a progressive learning strategy is proposed to optimize multi-resolution hash encodings from coarse to fine. In addition, we extend our method for reconstructing dynamic scenes with an incremental training strategy. Our experiments on various datasets demonstrate that NeuS2 significantly outperforms the state-of-the-arts in both surface reconstruction accuracy and training speed. The video is available at https://vcai.mpi-inf.mpg.de/projects/NeuS2/ .

神经辐射场（NERF）最近在新型视图合成中取得了令人印象深刻的结果。但是，以前的NERF作品主要关注以对象为中心的方案。在这项工作中，我们提出了360ROAM，这是一种新颖的场景级NERF系统，可以实时合成大型室内场景的图像并支持VR漫游。我们的系统首先从多个输入$ 360^\ circ $图像构建全向神经辐射场360NERF。然后，我们逐步估算一个3D概率的占用图，该概率占用图代表了空间密度形式的场景几何形状。跳过空的空间和上采样占据的体素本质上可以使我们通过以几何学意识的方式使用360NERF加速量渲染。此外，我们使用自适应划分和扭曲策略来减少和调整辐射场，以进一步改进。从占用地图中提取的场景的平面图可以为射线采样提供指导，并促进现实的漫游体验。为了显示我们系统的功效，我们在各种场景中收集了$ 360^\ Circ $图像数据集并进行广泛的实验。基线之间的定量和定性比较说明了我们在复杂室内场景的新型视图合成中的主要表现。

我们呈现高动态范围神经辐射字段（HDR-NERF），以从一组低动态范围（LDR）视图的HDR辐射率字段与不同的曝光。使用HDR-NERF，我们能够在不同的曝光下生成新的HDR视图和新型LDR视图。我们方法的关键是模拟物理成像过程，该过程决定了场景点的辐射与具有两个隐式功能的LDR图像中的像素值转换为：RADIACE字段和音调映射器。辐射场对场景辐射（值在0到+末端之间的值变化），其通过提供相应的射线源和光线方向来输出光线的密度和辐射。 TONE MAPPER模拟映射过程，即在相机传感器上击中的光线变为像素值。通过将辐射和相应的曝光时间送入音调映射器来预测光线的颜色。我们使用经典的卷渲染技术将输出辐射，颜色和密度投影为HDR和LDR图像，同时只使用输入的LDR图像作为监控。我们收集了一个新的前瞻性的HDR数据集，以评估所提出的方法。综合性和现实世界场景的实验结果验证了我们的方法不仅可以准确控制合成视图的曝光，还可以用高动态范围呈现视图。

We present a method that synthesizes novel views of complex scenes by interpolating a sparse set of nearby views. The core of our method is a network architecture that includes a multilayer perceptron and a ray transformer that estimates radiance and volume density at continuous 5D locations (3D spatial locations and 2D viewing directions), drawing appearance information on the fly from multiple source views. By drawing on source views at render time, our method hearkens back to classic work on image-based rendering (IBR), and allows us to render high-resolution imagery. Unlike neural scene representation work that optimizes per-scene functions for rendering, we learn a generic view interpolation function that generalizes to novel scenes. We render images using classic volume rendering, which is fully differentiable and allows us to train using only multiview posed images as supervision. Experiments show that our method outperforms recent novel view synthesis methods that also seek to generalize to novel scenes. Further, if fine-tuned on each scene, our method is competitive with state-of-the-art single-scene neural rendering methods. 1

我们向渲染和时间（4D）重建人类的渲染和时间（4D）重建的神经辐射场，通过稀疏的摄像机捕获或甚至来自单眼视频。我们的方法将思想与神经场景表示，新颖的综合合成和隐式统计几何人称的人类表示相结合，耦合使用新颖的损失功能。在先前使用符号距离功能表示的结构化隐式人体模型，而不是使用统一的占用率来学习具有统一占用的光域字段。这使我们能够从稀疏视图中稳健地融合信息，并概括超出在训练中观察到的姿势或视图。此外，我们应用几何限制以共同学习观察到的主题的结构 - 包括身体和衣服 - 并将辐射场正规化为几何合理的解决方案。在多个数据集上的广泛实验证明了我们方法的稳健性和准确性，其概括能力显着超出了一系列的姿势和视图，以及超出所观察到的形状的统计外推。

我们提出了一种可区分的渲染算法，以进行有效的新型视图合成。通过偏离基于音量的表示，支持学习点表示，我们在训练和推理方面的内存和运行时范围内改进了现有方法的数量级。该方法从均匀采样的随机点云开始，并使用基于可区分的SPLAT渲染器来发展模型以匹配一组输入图像，从而学习了每点位置和观看依赖性外观。在训练和推理中，我们的方法比NERF快300倍，质量只有边缘牺牲，而在静态场景中使用少于10 〜MB的记忆。对于动态场景，我们的方法比Stnerf训练两个数量级，并以接近互动速率渲染，同时即使在不施加任何时间固定的正则化合物的情况下保持较高的图像质量和时间连贯性。

Figure 1: Our method can synthesize novel views in both space and time from a single monocular video of a dynamic scene. Here we show video results with various configurations of fixing and interpolating view and time (left), as well as a visualization of the recovered scene geometry (right). Please view with Adobe Acrobat or KDE Okular to see animations.

最近的神经人类表示可以产生高质量的多视图渲染，但需要使用密集的多视图输入和昂贵的培训。因此，它们在很大程度上仅限于静态模型，因为每个帧都是不可行的。我们展示了人类学 - 一种普遍的神经表示 - 用于高保真自由观察动态人类的合成。类似于IBRNET如何通过避免每场景训练来帮助NERF，Humannerf跨多视图输入采用聚合像素对准特征，以及用于解决动态运动的姿势嵌入的非刚性变形场。原始人物员已经可以在稀疏视频输入的稀疏视频输入上产生合理的渲染。为了进一步提高渲染质量，我们使用外观混合模块增强了我们的解决方案，用于组合神经体积渲染和神经纹理混合的益处。各种多视图动态人类数据集的广泛实验证明了我们在挑战运动中合成照片 - 现实自由观点的方法和非常稀疏的相机视图输入中的普遍性和有效性。