Pattern images are everywhere in the digital and physical worlds, and tools to edit them are valuable. But editing pattern images is tricky: desired edits are often programmatic: structure-aware edits that alter the underlying program which generates the pattern. One could attempt to infer this underlying program, but current methods for doing so struggle with complex images and produce unorganized programs that make editing tedious. In this work, we introduce a novel approach to perform programmatic edits on pattern images. By using a pattern analogy—a pair of simple patterns to demonstrate the intended edit—and a learning-based generative model to execute these edits, our method allows users to intuitively edit patterns. To enable this paradigm, we introduce SplitWeave, a domain-specific language that, combined with a framework for sampling synthetic pattern analogies, enables the creation of a large, high-quality synthetic training dataset. We also present TriFuser, a Latent Diffusion Model (LDM) designed to overcome critical issues that arise when naively deploying LDMs to this task. Extensive experiments on real-world, artist-sourced patterns reveals that our method faithfully performs the demonstrated edit while also generalizing to related pattern styles beyond its training distribution.

Generative methods for image and video editing leverage generative models as priors to perform edits despite incomplete information, such as changing the composition of 3D objects depicted in a single image. Recent methods have shown promising composition editing results for images. However, in the video domain, editing methods have focused on editing objects’ appearance and motion, or camera motion. As a result, methods for editing object composition in videos remain largely unexplored. We propose VideoHandles as a method for editing 3D object compositions in videos of static scenes with camera motion. Our approach enables the editing of an object’s 3D position across all frames of a video in a temporally consistent manner. This is achieved by lifting intermediate features of a generative model to a 3D reconstruction that is shared between all frames, editing the reconstruction, and projecting the features on the edited reconstruction back to each frame. To the best of our knowledge, this is the first generative approach to edit object compositions in videos. Our approach is simple and training-free, while outperforming state-of-the-art image editing baselines. Our project page is https://videohandles.github.io.

Predicting diverse object motions from a single static image remains challenging, as current video generation models often entangle object movement with camera motion and other scene changes. While recent methods can predict specific motions from motion arrow input, they rely on synthetic data and predefined motions, limiting their application to complex scenes. We introduce Motion Modes, a training-free approach that explores a pre-trained imageto-video generator’s latent distribution to discover various distinct and plausible motions focused on selected objects in static images. We achieve this by employing a flow generator guided by energy functions designed to disentangle object and camera motion. Additionally, we use an energy inspired by particle guidance [8] to diversify the generated motions, without requiring explicit training data. Experimental results demonstrate that Motion Modes generates realistic and varied object animations, surpassing previous methods and even human predictions regarding plausibility and diversity.

Simple as it seems, moving an object to another location within an image is, in fact, a challenging image-editing task that requires re-harmonizing the lighting, adjusting the pose based on perspective, accurately filling occluded regions, and ensuring coherent synchronization of shadows and reflections while maintaining the object identity. In this paper, we present ObjectMover, a generative model that can perform object movement in highly challenging scenes. Our key insight is that we model this task as a sequence-to-sequence problem and fine-tune a video generation model to leverage its knowledge of consistent object generation across video frames. We show that with this approach, our model is able to adjust to complex real-world scenarios, handling extreme lighting harmonization and object effect movement. As large-scale data for object movement are unavailable, we construct a data generation pipeline using a modern game engine to synthesize high-quality data pairs. We further propose a multi-task learning strategy that enables training on real-world video data to improve the model generalization. Through extensive experiments, we demonstrate that ObjectMover achieves outstanding results and adapts well to real-world scenarios.

We train a feed-forward text-to-3D diffusion generator for human characters using only single-view 2D data for supervision. Existing 3D generative models cannot yet match the fidelity of image and/or video generative models. State-of-the-art 3D generators are either trained with explicit 3D supervision and are thus limited by the volume and diversity of existing 3D data. Meanwhile, generators that can be trained with only 2D data as supervision typically produce coarser results, cannot be text-conditioned, and/or must revert to test-time optimization. We observe that GAN- and diffusion-based generators have complementary qualities: GANs can be trained efficiently with 2D supervision to produce high-quality 3D objects but are hard to condition on text. In contrast, denoising diffusion models can be conditioned efficiently but tend to be hard to train with only 2D supervision. We introduce GANFusion that starts by generating unconditional triplane features for 3D data using a GAN architecture trained with only single-view 2D data. We then generate random samples from the GAN, caption them, and train a text-conditioned diffusion model that directly learns to sample from the space of good triplane features that can be decoded into 3D objects. We evaluate the proposed method in the context of text-conditioned full-body human generation and show improvements over possible alternatives.

We propose CLASS; a novel unified model for the syn-thesis and search for design layouts, two tasks that are often handled separately by prior works. We propose to learn a compact and coherent latent feature of a layout supporting joint search and synthesis. This allows vari-ous operations such style-conditioned layout generation, la-tent space manipulation and provides seamless integration of search and synthesis for an effective design workflow. We train CLASS with a dual decoder: a new transformer-based layout-conditioned decoder and a CNN-based raster decoder. The latent-conditioned decoder explicitly conditions upon a latent vector while generating a layout in an auto-regressive fashion. We train CLASS under variational framework which in conjunction with a raster-decoder en-hances the latent representation improving both generation and retrieval performances. We show the effectiveness of CLASS on the RICO and PubLayNet benchmarks, and demonstrate that CLASS is capable of high-quality synthe-sis from scratch, as well as performing self-completion, in-terpolation, project between design layouts, whilst achieving close to or better than state-of-the-art search performance.

Diffusion Handles is a novel approach to enable 3D object edits on diffusion images, requiring only existing pre-trained diffusion models depth estimation, without any fine-tuning or 3D object retrieval. The edited results remain plausible, photo-real, and preserve object identity. Diffusion Handles address a critically missing facet of generative image-based creative design. Our key insight is to lift diffusion activations for a selected object to 3D using a proxy depth, 3D-transform the depth and associated activations, and project them back to image space. The diffusion process guided by the manipulated activations produces plausible edited images showing complex 3D occlusion and lighting effects. We evaluate Diffusion Handles: quantitatively, on a large synthetic data benchmark; and qualitatively by a user study, showing our output to be more plausible, and better than prior art at both, 3D editing and identity control.

Abstract 3D surface reconstruction from point clouds is a key step in areas such as content creation, archaeology, digital cultural heritage and engineering. Current approaches either try to optimize a non‐data‐driven surface representation to fit the points, or learn a data‐driven prior over the distribution of commonly occurring surfaces and how they correlate with potentially noisy point clouds. Data‐driven methods enable robust handling of noise and typically either focus on a global or a local prior, which trade‐off between robustness to noise on the global end and surface detail preservation on the local end. We propose PPSurf as a method that combines a global prior based on point convolutions and a local prior based on processing local point cloud patches. We show that this approach is robust to noise while recovering surface details more accurately than the current state‐of‐the‐art. Our source code, pre‐trained model and dataset are available at https://github.com/cg‐tuwien/ppsurf .

Modern 3D content creation heavily relies on procedural assets. In particular, procedural materials are ubiquitous in the industry, but their manipulation remains challenging. Previous work [Hu et al. 2023] conditionally generates procedural graphs that match a given input image. However, the parameter generation step limits how accurately the generated graph matches the input image, due to a reliance on supervision with scarcely available procedural data. We propose to improve parameter prediction accuracy for image-conditioned procedural material generation by leveraging reinforcement learning (RL) and present the first RL approach for procedural materials. RL circumvents the limited availability of procedural data, the domain gap between real and synthetic materials, and the need for end-to-end differentiable loss functions. Given a target image, we retrieve a procedural material and use an RL-trained transformer model to predict a set of parameters that reconstruct the target image as closely as possible. We show that using RL significantly improves parameter prediction to match a given target image compared to supervised methods on both synthetic and real target images.

Diffusion models currently achieve state-of-the-art performance for both conditional and unconditional image generation. However, so far, image diffusion models do not support tasks required for 3D understanding, such as view-consistent 3D generation or single-view object reconstruction. In this paper, we present RenderDiffusion, the first diffusion model for 3D generation and inference, trained using only monocular 2D supervision. Central to our method is a novel image denoising architecture that generates and renders an intermediate three-dimensional representation of a scene in each denoising step. This enforces a strong inductive structure within the diffusion process, providing a 3D consistent representation while only requiring 2D supervision. The resulting 3D representation can be rendered from any view. We evaluate RenderDiffusion on FFHQ, AFHQ, ShapeNet and CLEVR datasets, showing competitive performance for generation of 3D scenes and inference of 3D scenes from 2D images. Additionally, our diffusion-based approach allows us to use 2D inpainting to edit 3D scenes.

Representing a 3D shape with a set of primitives can aid perception of structure, improve robotic object manipulation, and enable editing, stylization, and compression of 3D shapes. Existing methods either use simple parametric primitives or learn a generative shape space of parts. Both have limitations: parametric primitives lead to coarse approximations, while learned parts offer too little control over the decomposition. We instead propose to decompose shapes using a library of 3D parts provided by the user, giving full control over the choice of parts. The library can contain parts with high-quality geometry that are suitable for a given category, resulting in meaningful decompositions with clean geometry. The type of decomposition can also be controlled through the choice of parts in the library. Our method works via a unsupervised approach that iteratively retrieves parts from the library and refines their placements. We show that this approach gives higher reconstruction accuracy and more desirable decompositions than existing approaches. Additionally, we show how the decomposition can be controlled through the part library by using different part libraries to reconstruct the same shapes.

Exploring variations of 3D shapes is a time-consuming process in traditional 3D modeling tools. Deep generative models of 3D shapes often feature continuous latent spaces that can, in principle, be used to explore potential variations starting from a set of input shapes; in practice, doing so can be problematic—latent spaces are high dimensional and hard to visualize, contain shapes that are not relevant to the input shapes, and linear paths through them often lead to sub-optimal shape transitions. Furthermore, one would ideally be able to explore variations in the original high-quality meshes used to train the generative model, not its lower-quality output geometry. In this paper, we present a method to explore variations among a given set of landmark shapes by constructing a mapping from an easily-navigable 2D exploration space to a subspace of a pre-trained generative model. We first describe how to find a mapping that spans the set of input landmark shapes and exhibits smooth variations between them. We then show how to turn the variations in this subspace into deformation fields, to transfer those variations to high-quality meshes for the landmark shapes. Our results show that our method can produce visually-pleasing and easily-navigable 2D exploration spaces for several different shape categories, especially as compared to prior work on learning deformation spaces for 3D shapes.

Abstract Procedural models (i.e. symbolic programs that output visual data) are a historically‐popular method for representing graphics content: vegetation, buildings, textures, etc. They offer many advantages: interpretable design parameters, stochastic variations, high‐quality outputs, compact representation, and more. But they also have some limitations, such as the difficulty of authoring a procedural model from scratch. More recently, AI‐based methods, and especially neural networks, have become popular for creating graphic content. These techniques allow users to directly specify desired properties of the artifact they want to create (via examples, constraints, or objectives), while a search, optimization, or learning algorithm takes care of the details. However, this ease of use comes at a cost, as it's often hard to interpret or manipulate these representations. In this state‐of‐the‐art report, we summarize research on neurosymbolic models in computer graphics: methods that combine the strengths of both AI and symbolic programs to represent, generate, and manipulate visual data. We survey recent work applying these techniques to represent 2D shapes, 3D shapes, and materials & textures. Along the way, we situate each prior work in a unified design space for neurosymbolic models, which helps reveal underexplored areas and opportunities for future research.

Abstract Lightweight material capture methods require a material prior, defining the subspace of plausible textures within the large space of unconstrained texel grids. Previous work has either used deep neural networks (trained on large synthetic material datasets) or procedural node graphs (constructed by expert artists) as such priors. In this paper, we propose a semi‐procedural differentiable material prior that represents materials as a set of (typically procedural) grayscale noises and patterns that are processed by a sequence of lightweight learnable convolutional filter operations. We demonstrate that the restricted structure of this architecture acts as an inductive bias on the space of material appearances, allowing us to optimize the weights of the convolutions per‐material, with no need for pre‐training on a large dataset. Combined with a differentiable rendering step and a perceptual loss, we enable single‐image tileable material capture comparable with state of the art. Our approach does not target the pixel‐perfect recovery of the material, but rather uses noises and patterns as input to match the target appearance. To achieve this, it does not require complex procedural graphs, and has a much lower complexity, computational cost and storage cost. We also enable control over the results, through changing the provided patterns and using guide maps to push the material properties towards a user‐driven objective.

We introduce ShapeCoder, the first system capable of taking a dataset of shapes, represented with unstructured primitives, and jointly discovering (i) useful abstraction functions and (ii) programs that use these abstractions to explain the input shapes. The discovered abstractions capture common patterns (both structural and parametric) across a dataset, so that programs rewritten with these abstractions are more compact, and suppress spurious degrees of freedom. ShapeCoder improves upon previous abstraction discovery methods, finding better abstractions, for more complex inputs, under less stringent input assumptions. This is principally made possible by two methodological advancements: (a) a shape-to-program recognition network that learns to solve sub-problems and (b) the use of e-graphs, augmented with a conditional rewrite scheme, to determine when abstractions with complex parametric expressions can be applied, in a tractable manner. We evaluate ShapeCoder on multiple datasets of 3D shapes, where primitive decompositions are either parsed from manual annotations or produced by an unsupervised cuboid abstraction method. In all domains, ShapeCoder discovers a library of abstractions that captures high-level relationships, removes extraneous degrees of freedom, and achieves better dataset compression compared with alternative approaches. Finally, we investigate how programs rewritten to use discovered abstractions prove useful for downstream tasks.

This work is concerned with a representation of shapes that disentangles fine, local and possibly repeating geometry, from global, coarse structures. Achieving such disentanglement leads to two unrelated advantages: i) a significant compression in the number of parameters required to represent a given geometry; ii) the ability to manipulate either global geometry, or local details, without harming the other. At the core of our approach lies a novel pipeline and neural architecture, which are optimized to represent one specific atlas, representing one 3D surface. Our pipeline and architecture are designed so that disentanglement of global geometry from local details is accomplished through optimization, in a completely unsupervised manner. We show that this approach achieves better neural shape compression than the state of the art, as well as enabling manipulation and transfer of shape details. Project page http://geometry.cs.ucl.ac.uk/projects/2022/cnnmaps/.

Neural representations are popular for representing shapes as they can be used for data cleanup, model completion, shape editing, and shape synthesis. Current neural representations can be categorized as either overfitting to a single object instance, or representing a collection of objects. However, neither allows accurate editing of neural scene representations: on the one hand, methods that overfit objects achieve highly accurate reconstructions but do not support editing, as they do not generalize to unseen object configurations; on the other hand, methods that represent a family of objects with variations do generalize but produce approximate reconstructions. We propose NeuForm to combine the advantages of both overfitted and generalizable representations by adaptively overfitting a generalizable representation to regions where reliable data is available, while using the generalizable representation everywhere else. We achieve this with a carefully designed architecture and an approach that blends the network weights of the two representations. We demonstrate edits that successfully reconfigure parts of human-made shapes, such as chairs, tables, and lamps, while preserving the accuracy of an overfitted shape representation. We compare with two state-of-the-art competitors and demonstrate clear improvements in terms of plausibility and fidelity of the resultant edits.

We address the problem of indoor layout synthesis, which is a topic of continuing research interest in computer graphics. The newest works made significant progress using data-driven generative methods; however, these approaches rely on suitable datasets. In practice, desirable layout properties may not exist in a dataset, for instance, specific expert knowledge can be missing in the data. We propose a method that combines expert knowledge, for example, knowledge about ergonomics, with a data-driven generator based on the popular Transformer architecture. The knowledge is given as differentiable scalar functions, which can be used both as weights or as additional terms in the loss function. Using this knowledge, the synthesized layouts can be biased to exhibit desirable properties, even if these properties are not present in the dataset. Our approach can also alleviate problems of lack of data and imperfections in the data. Our work aims to improve generative machine learning for modeling and provide novel tools for designers and amateurs for the problem of interior layout creation.

Recent methods (e.g. MaterialGAN) have used unconditional GANs to generate per-pixel material maps, or as a prior to reconstruct materials from input photographs. These models can generate varied random material appearance, but do not have any mechanism to constrain the generated material to a specific category or to control the coarse structure of the generated material, such as the exact brick layout on a brick wall. Furthermore, materials reconstructed from a single input photo commonly have artifacts and are generally not tileable, which limits their use in practical content creation pipelines. We propose TileGen, a generative model for SVBRDFs that is specific to a material category, always tileable, and optionally conditional on a provided input structure pattern. TileGen is a variant of StyleGAN whose architecture is modified to always produce tileable (periodic) material maps. In addition to the standard “style” latent code, TileGen can optionally take a condition image, giving a user direct control over the dominant spatial (and optionally color) features of the material. For example, in brick materials, the user can specify a brick layout and the brick color, or in leather materials, the locations of wrinkles and folds. Our inverse rendering approach can find a material perceptually matching a single target photograph by optimization. This reconstruction can also be conditional on a user-provided pattern. The resulting materials are tileable, can be larger than the target image, and are editable by varying the condition.

Abstract Despite the ubiquitous use of materials maps in modern rendering pipelines, their editing and control remains a challenge. In this paper, we present an example‐based material control method to augment input material maps based on user‐provided material photos. We train a tileable version of MaterialGAN and leverage its material prior to guide the appearance transfer, optimizing its latent space using differentiable rendering. Our method transfers the micro and meso‐structure textures of user provided target(s) photographs, while preserving the structure and quality of the input material. We show our methods can control existing material maps, increasing realism or generating new, visually appealing materials.

Procedural material graphs are a compact, parameteric, and resolution-independent representation that are a popular choice for material authoring. However, designing procedural materials requires significant expertise and publicly accessible libraries contain only a few thousand such graphs. We present MatFormer, a generative model that can produce a diverse set of high-quality procedural materials with complex spatial patterns and appearance. While procedural materials can be modeled as directed (operation) graphs, they contain arbitrary numbers of heterogeneous nodes with unstructured, often long-range node connections, and functional constraints on node parameters and connections. MatFormer addresses these challenges with a multi-stage transformer-based model that sequentially generates nodes, node parameters, and edges, while ensuring the semantic validity of the graph. In addition to generation, MatFormer can be used for the auto-completion and exploration of partial material graphs. We qualitatively and quantitatively demonstrate that our method outperforms alternative approaches, in both generated graph and material quality.

We present a method for reconstructing triangle meshes from point clouds. Existing learning-based methods for mesh reconstruction mostly generate triangles individually, making it hard to create manifold meshes. We leverage the properties of 2D Delaunay triangulations to construct a mesh from manifold surface elements. Our method first estimates local geodesic neighborhoods around each point. We then perform a 2D projection of these neighborhoods using a learned logarithmic map. A Delaunay triangulation in this 2D domain is guaranteed to produce a manifold patch, which we call a Delaunay surface element. We synchronize the local 2D projections of neighboring elements to maximize the manifoldness of the reconstructed mesh. Our results show that we achieve better overall manifoldness of our reconstructed meshes than current methods to reconstruct meshes with arbitrary topology. Our code, data and pretrained models can be found online: https://github.com/mrakotosaon/dse-meshing

We propose a new generative model for layout generation. We generate layouts in three steps. First, we generate the layout elements as nodes in a layout graph. Second, we compute constraints between layout elements as edges in the layout graph. Third, we solve for the final layout using constrained optimization. For the first two steps, we build on recent transformer architectures. The layout optimization implements the constraints efficiently. We show three practical contributions compared to the state of the art: our work requires no user input, produces higher quality layouts, and enables many novel capabilities for conditional layout generation.

Computer-aided design (CAD) is the most widely used modeling approach for technical design. The typical starting point in these designs is 2D sketches which can later be extruded and combined to obtain complex three-dimensional assemblies. Such sketches are typically composed of parametric primitives, such as points, lines, and circular arcs, augmented with geometric constraints linking the primitives, such as coincidence, parallelism, or orthogonality. Sketches can be represented as graphs, with the primitives as nodes and the constraints as edges. Training a model to automatically generate CAD sketches can enable several novel workflows, but is challenging due to the complexity of the graphs and the heterogeneity of the primitives and constraints. In particular, each type of primitive and constraint may require a record of different size and parameter types.We propose SketchGen as a generative model based on a transformer architecture to address the heterogeneity problem by carefully designing a sequential language for the primitives and constraints that allows distinguishing between different primitive or constraint types and their parameters, while encouraging our model to re-use information across related parameters, encoding shared structure. A particular highlight of our work is the ability to produce primitives linked via constraints that enables the final output to be further regularized via a constraint solver. We evaluate our model by demonstrating constraint prediction for given sets of primitives and full sketch generation from scratch, showing that our approach significantly out performs the state-of-the-art in CAD sketch generation.

A popular way to create detailed yet easily controllable 3D shapes is via procedural modeling, i.e. generating geometry using programs. Such programs consist of a series of instructions along with their associated parameter values. To fully realize the benefits of this representation, a shape program should be compact and only expose degrees of freedom that allow for meaningful manipulation of output geometry. One way to achieve this goal is to design higher-level macro operators that, when executed, expand into a series of commands from the base shape modeling language. However, manually authoring such macros, much like shape programs themselves, is difficult and largely restricted to domain experts. In this paper, we present ShapeMOD, an algorithm for automatically discovering macros that are useful across large datasets of 3D shape programs. ShapeMOD operates on shape programs expressed in an imperative, statement-based language. It is designed to discover macros that make programs more compact by minimizing the number of function calls and free parameters required to represent an input shape collection. We run ShapeMOD on multiple collections of programs expressed in a domain-specific language for 3D shape structures. We show that it automatically discovers a concise set of macros that abstract out common structural and parametric patterns that generalize over large shape collections. We also demonstrate that the macros found by ShapeMOD improve performance on downstream tasks including shape generative modeling and inferring programs from point clouds. Finally, we conduct a user study that indicates that ShapeMOD's discovered macros make interactive shape editing more efficient.

Triangle meshes remain the most popular data representation for surface geometry. This ubiquitous representation is essentially a hybrid one that decouples continuous vertex locations from the discrete topological triangulation. Unfortunately, the combinatorial nature of the triangulation prevents taking derivatives over the space of possible meshings of any given surface. As a result, to date, mesh processing and optimization techniques have been unable to truly take advantage of modular gradient descent components of modern optimization frameworks. In this work, we present a differentiable surface triangulation that enables optimization for any per-vertex or per-face differentiable objective function over the space of underlying surface triangulations. Our method builds on the result that any 2D triangulation can be achieved by a suitably perturbed weighted Delaunay triangulation. We translate this result into a computational algorithm by proposing a soft relaxation of the classical weighted Delaunay triangulation and optimizing over vertex weights and vertex locations. We extend the algorithm to 3D by decomposing shapes into developable sets and differentiably meshing each set with suitable boundary constraints. We demonstrate the efficacy of our method on various planar and surface meshes on a range of difficult-to-optimize objective functions. Our code can be found online: https://github.com/mrakotosaon/diff-surface-triangulation.

Learning to encode differences in the geometry and (topological) structure of the shapes of ordinary objects is key to generating semantically plausible variations of a given shape, transferring edits from one shape to another, and for many other applications in 3D content creation. The common approach of encoding shapes as points in a high-dimensional latent feature space suggests treating shape differences as vectors in that space. Instead, we treat shape differences as primary objects in their own right and propose to encode them in their own latent space. In a setting where the shapes themselves are encoded in terms of fine-grained part hierarchies, we demonstrate that a separate encoding of shape deltas or differences provides a principled way to deal with inhomogeneities in the shape space due to different combinatorial part structures, while also allowing for compactness in the representation, as well as edit abstraction and transfer. Our approach is based on a conditional variational autoencoder for encoding and decoding shape deltas, conditioned on a source shape. We demonstrate the effectiveness and robustness of our approach in multiple shape modification and generation tasks, and provide comparison and ablation studies on the PartNet dataset, one of the largest publicly available 3D datasets.

Abstract Point clouds obtained with 3D scanners or by image‐based reconstruction techniques are often corrupted with significant amount of noise and outliers. Traditional methods for point cloud denoising largely rely on local surface fitting (e.g. jets or MLS surfaces), local or non‐local averaging or on statistical assumptions about the underlying noise model. In contrast, we develop a simple data‐driven method for removing outliers and reducing noise in unordered point clouds. We base our approach on a deep learning architecture adapted from PCPNet, which was recently proposed for estimating local 3D shape properties in point clouds. Our method first classifies and discards outlier samples, and then estimates correction vectors that project noisy points onto the original clean surfaces. The approach is efficient and robust to varying amounts of noise and outliers, while being able to handle large densely sampled point clouds. In our extensive evaluation, both on synthetic and real data, we show an increased robustness to strong noise levels compared to various state‐of‐the‐art methods, enabling accurate surface reconstruction from extremely noisy real data obtained by range scans. Finally, the simplicity and universality of our approach makes it very easy to integrate in any existing geometry processing pipeline. Both the code and pre‐trained networks can be found on the project page ( https://github.com/mrakotosaon/pointcleannet ).

Manually authoring 3D shapes is difficult and time consuming; generative models of 3D shapes offer compelling alternatives. Procedural representations are one such possibility: they offer high-quality and editable results but are difficult to author and often produce outputs with limited diversity. On the other extreme are deep generative models: given enough data, they can learn to generate any class of shape but their outputs have artifacts and the representation is not editable. In this paper, we take a step towards achieving the best of both worlds for novel 3D shape synthesis. First, we propose ShapeAssembly, a domain-specific "assembly-language" for 3D shape structures. ShapeAssembly programs construct shape structures by declaring cuboid part proxies and attaching them to one another, in a hierarchical and symmetrical fashion. ShapeAssembly functions are parameterized with continuous free variables, so that one program structure is able to capture a family of related shapes. We show how to extract ShapeAssembly programs from existing shape structures in the PartNet dataset. Then, we train a deep generative model, a hierarchical sequence VAE, that learns to write novel ShapeAssembly programs. Our approach leverages the strengths of each representation: the program captures the subset of shape variability that is interpretable and editable, and the deep generative model captures variability and correlations across shape collections that is hard to express procedurally. We evaluate our approach by comparing the shapes output by our generated programs to those from other recent shape structure synthesis models. We find that our generated shapes are more plausible and physically-valid than those of other methods. Additionally, we assess the latent spaces of these models, and find that ours is better structured and produces smoother interpolations. As an application, we use our generative model and differentiable program interpreter to infer and fit shape programs to unstructured geometry, such as point clouds.

Patterns, which are collections of elements arranged in regular or near-regular arrangements, are an important graphic art form and widely used due to their elegant simplicity and aesthetic appeal. When a pattern is encoded as a flat image without the underlying structure, manually editing the pattern is tedious and challenging as one has to both preserve the individual element shapes and their original relative arrangements. State-of-the-art deep learning frameworks that operate at the pixel level are unsuitable for manipulating such patterns. Specifically, these methods can easily disturb the shapes of the individual elements or their arrangement, and thus fail to preserve the latent structures of the input patterns. We present a novel differentiable compositing operator using pattern elements and use it to discover structures, in the form of a layered representation of graphical objects, directly from raw pattern images. This operator allows us to adapt current deep learning based image methods to effectively handle patterns. We evaluate our method on a range of patterns and demonstrate superiority in the context of pattern manipulations when compared against state-of-the-art pixel- or point-based alternatives.

The ability to generate novel, diverse, and realistic 3D shapes along with associated part semantics and structure is central to many applications requiring high-quality 3D assets or large volumes of realistic training data. A key challenge towards this goal is how to accommodate diverse shape variations, including both continuous deformations of parts as well as structural or discrete alterations which add to, remove from, or modify the shape constituents and compositional structure. Such object structure can typically be organized into a hierarchy of constituent object parts and relationships, represented as a hierarchy of n -ary graphs. We introduce StructureNet, a hierarchical graph network which (i) can directly encode shapes represented as such n -ary graphs, (ii) can be robustly trained on large and complex shape families, and (iii) be used to generate a great diversity of realistic structured shape geometries. Technically, we accomplish this by drawing inspiration from recent advances in graph neural networks to propose an order-invariant encoding of n -ary graphs, considering jointly both part geometry and inter-part relations during network training. We extensively evaluate the quality of the learned latent spaces for various shape families and show significant advantages over baseline and competing methods. The learned latent spaces enable several structure-aware geometry processing applications, including shape generation and interpolation, shape editing, or shape structure discovery directly from un-annotated images, point clouds, or partial scans.

Next generation smart and augmented reality systems demand a computational understanding of monocular footage that captures humans in physical spaces to reveal plausible object arrangements and human-object interactions. Despite recent advances, both in scene layout and human motion analysis, the above setting remains challenging to analyze due to regular occlusions that occur between objects and human motions. We observe that the interaction between object arrangements and human actions is often strongly correlated, and hence can be used to help recover from these occlusions. We present iMapper, a data-driven method to identify such human-object interactions and utilize them to infer layouts of occluded objects. Starting from a monocular video with detected 2D human joint positions that are potentially noisy and occluded, we first introduce the notion of interaction-saliency as space-time snapshots where informative human-object interactions happen. Then, we propose a global optimization to retrieve and fit interactions from a database to the detected salient interactions in order to best explain the input video. We extensively evaluate the approach, both quantitatively against manually annotated ground truth and through a user study, and demonstrate that iMapper produces plausible scene layouts for scenes with medium to heavy occlusion. Code and data are available on the project page.

Abstract In this paper, we propose PCPNET, a deep‐learning based approach for estimating local 3D shape properties in point clouds. In contrast to the majority of prior techniques that concentrate on global or mid‐level attributes, e.g., for shape classification or semantic labeling, we suggest a patch‐based learning method, in which a series of local patches at multiple scales around each point is encoded in a structured manner. Our approach is especially well‐adapted for estimating local shape properties such as normals (both unoriented and oriented) and curvature from raw point clouds in the presence of strong noise and multi‐scale features. Our main contributions include both a novel multi‐scale variant of the recently proposed PointNet architecture with emphasis on local shape information, and a series of novel applications in which we demonstrate how learning from training data arising from well‐structured triangle meshes, and applying the trained model to noisy point clouds can produce superior results compared to specialized state‐of‐the‐art techniques. Finally, we demonstrate the utility of our approach in the context of shape reconstruction, by showing how it can be used to extract normal orientation information from point clouds.

Coarse building mass models are now routinely generated at scales ranging from individual buildings to whole cities. Such models can be abstracted from raw measurements, generated procedurally, or created manually. However, these models typically lack any meaningful geometric or texture details, making them unsuitable for direct display. We introduce the problem of automatically and realistically decorating such models by adding semantically consistent geometric details and textures. Building on the recent success of generative adversarial networks (GANs), we propose F ranken GAN, a cascade of GANs that creates plausible details across multiple scales over large neighborhoods. The various GANs are synchronized to produce consistent style distributions over buildings and neighborhoods. We provide the user with direct control over the variability of the output. We allow him/her to interactively specify the style via images and manipulate style-adapted sliders to control style variability. We test our system on several large-scale examples. The generated outputs are qualitatively evaluated via a set of perceptual studies and are found to be realistic, semantically plausible, and consistent in style.

Finding similar points in globally or locally similar shapes has been studied extensively through the use of various point descriptors or shape-matching methods. However, little work exists on finding similar points in dissimilar shapes. In this paper, we present the results of a study where users were given two dissimilar two-dimensional shapes and asked to map a given point in the first shape to the point in the second shape they consider most similar. We find that user mappings in this study correlate strongly with simple geometric relationships between points and shapes. To predict the probability distribution of user mappings between any pair of simple two-dimensional shapes, two distinct statistical models are defined using these relationships. We perform a thorough validation of the accuracy of these predictions and compare our models qualitatively and quantitatively to well-known shape-matching methods. Using our predictive models, we propose an approach to map objects or procedural content between different shapes in different design scenarios.

In the context of scene understanding, a variety of methods exists to estimate different information channels from mono or stereo images, including disparity, depth, and normals. Although several advances have been reported in the recent years for these tasks, the estimated information is often imprecise particularly near depth discontinuities or creases. Studies have however shown that precisely such depth edges carry critical cues for the perception of shape, and play important roles in tasks like depth-based segmentation or foreground selection. Unfortunately, the currently extracted channels often carry conflicting signals, making it difficult for subsequent applications to effectively use them. In this paper, we focus on the problem of obtaining high-precision depth edges (i.e., depth contours and creases) by jointly analyzing such unreliable information channels. We propose DepthCut , a data-driven fusion of the channels using a convolutional neural network trained on a large dataset with known depth. The resulting depth edges can be used for segmentation, decomposing a scene into depth layers with relatively flat depth, or improving the accuracy of the depth estimate near depth edges by constraining its gradients to agree with these edges. Quantitatively, we compare against 18 variants of baselines and demonstrate that our depth edges result in an improved segmentation performance and an improved depth estimate near depth edges compared to data-agnostic channel fusion. Qualitatively, we demonstrate that the depth edges result in superior segmentation and depth orderings. (Code and datasets will be made available.)

Patterns play a central role in 2D graphic design. A critical step in the design of patterns is evaluating multiple design alternatives. Exploring these alternatives with existing tools is challenging because most tools force users to work with a single fixed representation of the pattern that encodes a specific set of geometric relationships between pattern elements. However, for most patterns, there are many different interpretations of its regularity that correspond to different design variations. The exponential nature of this variation space makes the problem of finding all variations intractable. We present a method called PATEX to characterize and efficiently identify distinct and valid pattern variations, allowing users to directly navigate the variation space. Technically, we propose a novel linear approximation to handle the complexity of the problem and efficiently enumerate suitable pattern variations under proposed element movements. We also present two pattern editing interfaces that expose the detected pattern variations as suggested edits to the user. We show a diverse collection of pattern edits and variations created with PATEX. The results from our user study indicate that our suggested variations can be useful and inspirational for typical pattern editing tasks.

As humans, we regularly interpret scenes based on how objects arerelated, rather than based on the objects themselves. For example, we see a personridingan object X or a plankbridgingtwo objects. Current methods provide limited support to search for content based on such relations. We presentraid, a relation-augmented image descriptor that supports queries based on inter-region relations. The key idea of our descriptor is to encode region-to-region relations as the spatial distribution of point-to-region relationships between two image regions.raidallows sketch-based retrieval and requires minimal training data, thus making it suited even for querying uncommon relations. We evaluate the proposed descriptor by querying into large image databases and successfully extract non-trivial images demonstrating complex inter-region relations, which are easily missed or erroneously classified by existing methods. We assess the robustness ofraidon multiple datasets even when the region segmentation is computed automatically or very noisy.

We propose a novel example-based approach to synthesize scenes with complex relations, e.g., when one object is 'hooked', 'surrounded', 'contained' or 'tucked into' another object. Existing relationship descriptors used in automatic scene synthesis methods are based on contacts or relative vectors connecting the object centers. Such descriptors do not fully capture the geometry of spatial interactions, and therefore cannot describe complex relationships. Our idea is to enrich the description of spatial relations between object surfaces by encoding the geometry of the open space around objects, and use this as a template for fitting novel objects. To this end, we introduce relationship templates as descriptors of complex relationships; they are computed from an example scene and combine the interaction bisector surface (IBS) with a novel feature called the space coverage feature (SCF), which encodes the open space in the frequency domain. New variations of a scene can be synthesized efficiently by fitting novel objects to the template. Our method greatly enhances existing automatic scene synthesis approaches by allowing them to handle complex relationships, as validated by our user studies. The proposed method generalizes well, as it can form complex relationships with objects that have a topology and geometry very different from the example scene.

Abstract In this paper, we present a method for non‐rigid, partial shape matching in vector graphics. Given a user‐specified query region in a 2D shape, similar regions are found, even if they are non‐linearly distorted. Furthermore, a non‐linear mapping is established between the query regions and these matches, which allows the automatic transfer of editing operations such as texturing. This is achieved by a two‐step approach. First, pointwise correspondences between the query region and the whole shape are established. The transformation parameters of these correspondences are registered in an appropriate transformation space. For transformations between similar regions, these parameters form surfaces in transformation space, which are extracted in the second step of our method. The extracted regions may be related to the query region by a non‐rigid transform, enabling non‐rigid shape matching.

We present a method to learn and propagate shape placements in 2D polygonal scenes from a few examples provided by a user. The placement of a shape is modeled as an oriented bounding box. Simple geometric relationships between this bounding box and nearby scene polygons define a feature set for the placement. The feature sets of all example placements are then used to learn a probabilistic model over all possible placements and scenes. With this model, we can generate a new set of placements with similar geometric relationships in any given scene. We introduce extensions that enable propagation and generation of shapes in 3D scenes, as well as the application of a learned modeling session to large scenes without additional user interaction. These concepts allow us to generate complex scenes with thousands of objects with relatively little user interaction.

We propose a method for propagating edit operations in 2D vector graphics, based on geometric relationship functions. These functions quantify the geometric relationship of a point to a polygon, such as the distance to the boundary or the direction to the closest corner vertex. The level sets of the relationship functions describe points with the same relationship to a polygon. For a given query point, we first determine a set of relationships to local features, construct all level sets for these relationships, and accumulate them. The maxima of the resulting distribution are points with similar geometric relationships. We show extensions to handle mirror symmetries, and discuss the use of relationship functions as local coordinate systems. Our method can be applied, for example, to interactive floorplan editing, and it is especially useful for large layouts, where individual edits would be cumbersome. We demonstrate populating 2D layouts with tens to hundreds of objects by propagating relatively few edit operations.

Abstract We present a method for rendering approximate soft shadows and diffuse indirect illumination in dynamic scenes. The proposed method approximates the original scene geometry with a set of tightly fitting spheres. In previous work, such spheres have been used to dynamically evaluate the visibility function to render soft shadows. In this paper, each sphere also acts as a low‐frequency secondary light source, thereby providing diffuse one‐bounce indirect illumination. The method is completely dynamic and proceeds in two passes: In a first pass, the light intensity distribution on each sphere is updated based on sample points on the corresponding object surface and converted into the spherical harmonics basis. In a second pass, this radiance information and the visibility are accumulated to shade final image pixels. The sphere approximation allows us to compute visibility and diffuse reflections of an object at interactive frame rates of over 20 fps for moderately complex scenes.