For modeling the 3D world behind 2D images, which 3D representation is most appropriate? A polygon mesh is a promising candidate for its compactness and geometric properties. However, it is not straightforward to model a polygon mesh from 2D images using neural networks because the conversion from a mesh to an image, or rendering, involves a discrete operation called rasterization, which prevents back-propagation. Therefore, in this work, we propose an approximate gradient for rasterization that enables the integration of rendering into neural networks. Using this renderer, we perform single-image 3D mesh reconstruction with silhouette image supervision and our system outperforms the existing voxel-based approach. Additionally, we perform gradient-based 3D mesh editing operations, such as 2D-to-3D style transfer and 3D DeepDream, with 2D supervision for the first time. These applications demonstrate the potential of the integration of a mesh renderer into neural networks and the effectiveness of our proposed renderer.

Rendering bridges the gap between 2D vision and 3D scenes by simulating the physical process of image formation. By inverting such renderer, one can think of a learning approach to infer 3D information from 2D images. However, standard graphics renderers involve a fundamental discretization step called rasterization, which prevents the rendering process to be differentiable, hence able to be learned. Unlike the state-of-the-art differentiable renderers [29,19], which only approximate the rendering gradient in the back propagation, we propose a truly differentiable rendering framework that is able to (1) directly render colorized mesh using differentiable functions and (2) back-propagate efficient supervision signals to mesh vertices and their attributes from various forms of image representations, including silhouette, shading and color images. The key to our framework is a novel formulation that views rendering as an aggregation function that fuses the probabilistic contributions of all mesh triangles with respect to the rendered pixels. Such formulation enables our framework to flow gradients to the occluded and far-range vertices, which cannot be achieved by the previous state-of-thearts. We show that by using the proposed renderer, one can achieve significant improvement in 3D unsupervised singleview reconstruction both qualitatively and quantitatively. Experiments also demonstrate that our approach is able to handle the challenging tasks in image-based shape fitting, which remain nontrivial to existing differentiable renderers. Code is available at https://github.com/ ShichenLiu/SoftRas.

综合照片 - 现实图像和视频是计算机图形的核心，并且是几十年的研究焦点。传统上，使用渲染算法（如光栅化或射线跟踪）生成场景的合成图像，其将几何形状和材料属性的表示为输入。统称，这些输入定义了实际场景和呈现的内容，并且被称为场景表示（其中场景由一个或多个对象组成）。示例场景表示是具有附带纹理的三角形网格（例如，由艺术家创建），点云（例如，来自深度传感器），体积网格（例如，来自CT扫描）或隐式曲面函数（例如，截短的符号距离）字段）。使用可分辨率渲染损耗的观察结果的这种场景表示的重建被称为逆图形或反向渲染。神经渲染密切相关，并将思想与经典计算机图形和机器学习中的思想相结合，以创建用于合成来自真实观察图像的图像的算法。神经渲染是朝向合成照片现实图像和视频内容的目标的跨越。近年来，我们通过数百个出版物显示了这一领域的巨大进展，这些出版物显示了将被动组件注入渲染管道的不同方式。这种最先进的神经渲染进步的报告侧重于将经典渲染原则与学习的3D场景表示结合的方法，通常现在被称为神经场景表示。这些方法的一个关键优势在于它们是通过设计的3D-一致，使诸如新颖的视点合成捕获场景的应用。除了处理静态场景的方法外，我们还涵盖了用于建模非刚性变形对象的神经场景表示...

We propose a differentiable sphere tracing algorithm to bridge the gap between inverse graphics methods and the recently proposed deep learning based implicit signed distance function. Due to the nature of the implicit function, the rendering process requires tremendous function queries, which is particularly problematic when the function is represented as a neural network. We optimize both the forward and backward passes of our rendering layer to make it run efficiently with affordable memory consumption on a commodity graphics card. Our rendering method is fully differentiable such that losses can be directly computed on the rendered 2D observations, and the gradients can be propagated backwards to optimize the 3D geometry. We show that our rendering method can effectively reconstruct accurate 3D shapes from various inputs, such as sparse depth and multi-view images, through inverse optimization. With the geometry based reasoning, our 3D shape prediction methods show excellent generalization capability and robustness against various noises. * Work done while Shaohui Liu was an academic guest at ETH Zurich.

Generative models, as an important family of statistical modeling, target learning the observed data distribution via generating new instances. Along with the rise of neural networks, deep generative models, such as variational autoencoders (VAEs) and generative adversarial network (GANs), have made tremendous progress in 2D image synthesis. Recently, researchers switch their attentions from the 2D space to the 3D space considering that 3D data better aligns with our physical world and hence enjoys great potential in practice. However, unlike a 2D image, which owns an efficient representation (i.e., pixel grid) by nature, representing 3D data could face far more challenges. Concretely, we would expect an ideal 3D representation to be capable enough to model shapes and appearances in details, and to be highly efficient so as to model high-resolution data with fast speed and low memory cost. However, existing 3D representations, such as point clouds, meshes, and recent neural fields, usually fail to meet the above requirements simultaneously. In this survey, we make a thorough review of the development of 3D generation, including 3D shape generation and 3D-aware image synthesis, from the perspectives of both algorithms and more importantly representations. We hope that our discussion could help the community track the evolution of this field and further spark some innovative ideas to advance this challenging task.

Rapid advances in 2D perception have led to systems that accurately detect objects in real-world images. However, these systems make predictions in 2D, ignoring the 3D structure of the world. Concurrently, advances in 3D shape prediction have mostly focused on synthetic benchmarks and isolated objects. We unify advances in these two areas. We propose a system that detects objects in real-world images and produces a triangle mesh giving the full 3D shape of each detected object. Our system, called Mesh R-CNN, augments Mask R-CNN with a mesh prediction branch that outputs meshes with varying topological structure by first predicting coarse voxel representations which are converted to meshes and refined with a graph convolution network operating over the mesh's vertices and edges. We validate our mesh prediction branch on ShapeNet, where we outperform prior work on single-image shape prediction. We then deploy our full Mesh R-CNN system on Pix3D, where we jointly detect objects and predict their 3D shapes. Project page: https://gkioxari.github.io/meshrcnn/.

We propose an analysis-by-synthesis method for fast multi-view 3D reconstruction of opaque objects with arbitrary materials and illumination. State-of-the-art methods use both neural surface representations and neural rendering. While flexible, neural surface representations are a significant bottleneck in optimization runtime. Instead, we represent surfaces as triangle meshes and build a differentiable rendering pipeline around triangle rasterization and neural shading. The renderer is used in a gradient descent optimization where both a triangle mesh and a neural shader are jointly optimized to reproduce the multi-view images. We evaluate our method on a public 3D reconstruction dataset and show that it can match the reconstruction accuracy of traditional baselines and neural approaches while surpassing them in optimization runtime. Additionally, we investigate the shader and find that it learns an interpretable representation of appearance, enabling applications such as 3D material editing.

我们介绍了Amazon Berkeley对象（ABO），这是一个新的大型数据集，旨在帮助弥合真实和虚拟3D世界之间的差距。ABO包含产品目录图像，元数据和艺术家创建的3D模型，具有复杂的几何形状和与真实的家用物体相对应的物理基础材料。我们得出了具有挑战性的基准，这些基准利用ABO的独特属性，并测量最先进的对象在三个开放问题上的最新限制，以了解实际3D对象：单视3D 3D重建，材料估计和跨域多视图对象检索。

随着几个行业正在朝着建模大规模的3D虚拟世界迈进，因此需要根据3D内容的数量，质量和多样性来扩展的内容创建工具的需求变得显而易见。在我们的工作中，我们旨在训练Parterant 3D生成模型，以合成纹理网格，可以通过3D渲染引擎直接消耗，因此立即在下游应用中使用。 3D生成建模的先前工作要么缺少几何细节，因此在它们可以生成的网格拓扑中受到限制，通常不支持纹理，或者在合成过程中使用神经渲染器，这使得它们在常见的3D软件中使用。在这项工作中，我们介绍了GET3D，这是一种生成模型，该模型直接生成具有复杂拓扑，丰富几何细节和高保真纹理的显式纹理3D网格。我们在可区分的表面建模，可区分渲染以及2D生成对抗网络中桥接了最新成功，以从2D图像集合中训练我们的模型。 GET3D能够生成高质量的3D纹理网格，从汽车，椅子，动物，摩托车和人类角色到建筑物，对以前的方法进行了重大改进。

虚拟内容创建和互动在现代3D应用中起着重要作用，例如AR和VR。从真实场景中恢复详细的3D模型可以显着扩大其应用程序的范围，并在计算机视觉和计算机图形社区中进行了数十年的研究。我们提出了基于体素的隐式表面表示Vox-Surf。我们的Vox-Surf将空间分为有限的体素。每个体素将几何形状和外观信息存储在其角顶点。 Vox-Surf得益于从体素表示继承的稀疏性，几乎适用于任何情况，并且可以轻松地从多个视图图像中训练。我们利用渐进式训练程序逐渐提取重要体素，以进一步优化，以便仅保留有效的体素，从而大大减少了采样点的数量并增加了渲染速度。细素还可以视为碰撞检测的边界量。该实验表明，与其他方法相比，Vox-Surf表示可以学习精致的表面细节和准确的颜色，并以更少的记忆力和更快的渲染速度来学习。我们还表明，Vox-Surf在场景编辑和AR应用中可能更实用。

我们提出了一种有效的方法，用于从多视图图像观察中联合优化拓扑，材料和照明。与最近的多视图重建方法不同，通常在神经网络中产生纠缠的3D表示，我们将三角形网格输出具有空间不同的材料和环境照明，这些方法可以在任何传统的图形引擎中未修改。我们利用近期工作在可差异化的渲染中，基于坐标的网络紧凑地代表体积纹理，以及可微分的游行四边形，以便直接在表面网上直接实现基于梯度的优化。最后，我们介绍了环境照明的分流和近似的可分辨率配方，以有效地回收全频照明。实验表明我们的提取模型用于高级场景编辑，材料分解和高质量的视图插值，全部以三角形的渲染器（光栅化器和路径示踪剂）的交互式速率运行。

可微分的渲染是现代视觉中的重要操作，允许在现代机器学习框架中使用逆图形方法3D理解。显式形状表示（体素，点云或网格），而相对容易呈现，通常遭受有限的几何保真度或拓扑限制。另一方面，隐式表示（占用，距离或辐射字段）保持更大的保真度，但遭受复杂或低效的渲染过程，限制可扩展性。在这项工作中，我们努力解决具有新颖形状表示的缺点，允许在隐式架构内快速可分辨地渲染。构建隐式距离表示，我们定义了指向距离字段（DDF），将定向点（位置和方向）映射到表面可见性和深度。这种场可以通过网络衍生物能够使差分表面几何提取（例如，表面法线和曲率）能够容易地构成，并且允许提取经典无符号距离场。使用概率DDFS（PDDFS），我们展示了如何模拟底层字段中固有的不连续性。最后，我们将方法应用于拟合单一形状，未配对的3D感知生成图像建模和单像3D重建任务，通过我们表示的多功能性展示具有简单架构组件的强大性能。

单视图3D对象重建是一项基本且具有挑战性的计算机视觉任务，旨在从单视RGB图像中恢复3D形状。大多数现有的基于深度学习的重建方法都是​​在同一类别上培训和评估的，并且在处理训练过程中未见的新颖类别的物体时，它们无法正常工作。本文着眼于这个问题，解决了零照片的单视3D网格重建，以研究对看不见类别的模型概括，并鼓励模型从字面上重建对象。具体而言，我们建议一个端到端的两阶段网络Zeromesh，以打破重建中的类别边界。首先，我们将复杂的图像到网格映射分解为两个较简单的映射，即图像对点映射和点对点映射，而后者主要是几何问题，而不是对象类别的依赖。其次，我们在2D和3D特征空间中设计了局部特征采样策略，以捕获跨对象共享的局部几何形状，以增强模型概括。第三，除了传统的点对点监督外，我们还引入了多视图轮廓损失以监督表面生成过程，该过程提供了其他正则化，并进一步缓解了过度拟合的问题。实验结果表明，我们的方法在不同方案和各种指标下，特别是对于新颖对象而言，在Shapenet和Pix3D上的现有作品显着优于Shapenet和Pix3D的现有作品。

Learning-based 3D reconstruction methods have shown impressive results. However, most methods require 3D supervision which is often hard to obtain for real-world datasets. Recently, several works have proposed differentiable rendering techniques to train reconstruction models from RGB images. Unfortunately, these approaches are currently restricted to voxel-and mesh-based representations, suffering from discretization or low resolution. In this work, we propose a differentiable rendering formulation for implicit shape and texture representations. Implicit representations have recently gained popularity as they represent shape and texture continuously. Our key insight is that depth gradients can be derived analytically using the concept of implicit differentiation. This allows us to learn implicit shape and texture representations directly from RGB images. We experimentally show that our singleview reconstructions rival those learned with full 3D supervision. Moreover, we find that our method can be used for multi-view 3D reconstruction, directly resulting in watertight meshes.

In recent years, substantial progress has been achieved in learning-based reconstruction of 3D objects. At the same time, generative models were proposed that can generate highly realistic images. However, despite this success in these closely related tasks, texture reconstruction of 3D objects has received little attention from the research community and state-of-the-art methods are either limited to comparably low resolution or constrained experimental setups. A major reason for these limitations is that common representations of texture are inefficient or hard to interface for modern deep learning techniques. In this paper, we propose Texture Fields, a novel texture representation which is based on regressing a continuous 3D function parameterized with a neural network. Our approach circumvents limiting factors like shape discretization and parameterization, as the proposed texture representation is independent of the shape representation of the 3D object. We show that Texture Fields are able to represent high frequency texture and naturally blend with modern deep learning techniques. Experimentally, we find that Texture Fields compare favorably to state-of-the-art methods for conditional texture reconstruction of 3D objects and enable learning of probabilistic generative models for texturing unseen 3D models. We believe that Texture Fields will become an important building block for the next generation of generative 3D models.

Intelligent mesh generation (IMG) refers to a technique to generate mesh by machine learning, which is a relatively new and promising research field. Within its short life span, IMG has greatly expanded the generalizability and practicality of mesh generation techniques and brought many breakthroughs and potential possibilities for mesh generation. However, there is a lack of surveys focusing on IMG methods covering recent works. In this paper, we are committed to a systematic and comprehensive survey describing the contemporary IMG landscape. Focusing on 110 preliminary IMG methods, we conducted an in-depth analysis and evaluation from multiple perspectives, including the core technique and application scope of the algorithm, agent learning goals, data types, targeting challenges, advantages and limitations. With the aim of literature collection and classification based on content extraction, we propose three different taxonomies from three views of key technique, output mesh unit element, and applicable input data types. Finally, we highlight some promising future research directions and challenges in IMG. To maximize the convenience of readers, a project page of IMG is provided at \url{https://github.com/xzb030/IMG_Survey}.

where the highest resolution is required, using facial performance capture as a case in point.

Implicit shape representations, such as Level Sets, provide a very elegant formulation for performing computations involving curves and surfaces. However, including implicit representations into canonical Neural Network formulations is far from straightforward. This has consequently restricted existing approaches to shape inference, to significantly less effective representations, perhaps most commonly voxels occupancy maps or sparse point clouds.To overcome this limitation we propose a novel formulation that permits the use of implicit representations of curves and surfaces, of arbitrary topology, as individual layers in Neural Network architectures with end-to-end trainability. Specifically, we propose to represent the output as an oriented level set of a continuous and discretised embedding function. We investigate the benefits of our approach on the task of 3D shape prediction from a single image and demonstrate its ability to produce a more accurate reconstruction compared to voxel-based representations. We further show that our model is flexible and can be applied to a variety of shape inference problems.

现有3D网格模型的新型纹理合成是迈向现有模拟器的照片现实资产产生的重要一步。但是现有方法固有地在2D图像空间中起作用，这是从给定的摄像头的角度来看3D空间的投影。这些方法采用摄像头角度，3D模型信息，照明信息并生成逼真的2D图像。为了从另一个角度或照明产生一个逼真的图像，我们需要每次更改参数时进行计算上昂贵的远程通过。同样，很难为可以满足时间约束的模拟器生成此类图像，图像的序列应相似，但只需要根据需要更改照明的观点。该解决方案不能直接与搅拌机和虚幻引擎等现有工具集成。手动解决方案是昂贵且耗时的。因此，我们提出了一个称为Graph生成对抗网络（GGAN）的新系统，该系统可以生成纹理，可以将其直接集成到给定的3D网格模型中，该模型使用Blender和Unreal Engine之类的工具，可以轻松地从任何角度和照明条件进行模拟。

在视觉计算中，3D几何形状以许多不同的形式表示，包括网格，点云，体素电网，水平集和深度图像。每个表示都适用于不同的任务，从而使一个表示形式转换为另一个表示（前向地图）是一个重要且常见的问题。我们提出了全向距离字段（ODF），这是一种新的3D形状表示形式，该表示通过将深度从任何观看方向从任何3D位置存储到对象的表面来编码几何形状。由于射线是ODF的基本单元，因此可以轻松地从通用的3D表示和点云等常见的3D表示。与限制代表封闭表面的水平集方法不同，ODF是未签名的，因此可以对开放表面进行建模（例如服装）。我们证明，尽管在遮挡边界处存在固有的不连续性，但可以通过神经网络（Neururodf）有效地学习ODF。我们还引入了有效的前向映射算法，以转换odf to＆从常见的3D表示。具体而言，我们引入了一种有效的跳跃立方体算法，用于从ODF生成网格。实验表明，神经模型可以通过过度拟合单个对象学会学会捕获高质量的形状，并学会概括对共同的形状类别。