We introduce ShapeWords1, an approach for synthesizing images based on 3D shape guidance and text prompts. ShapeWords incorporates target 3D shape information within specialized tokens embedded together with the input text, effectively blending 3D shape awareness with textual context to guide the image synthesis process. Unlike conventional shape guidance methods that rely on depth maps restricted to fixed viewpoints and often overlook full 3D structure or textual context, ShapeWords generates diverse yet consistent images that reflect both the target shape’s geometry and the textual description. Experimental results show that ShapeWords produces images that are more text-compliant, aesthetically plausible, while also maintaining 3D shape awareness.

Extraction of building footprint polygons from remotely sensed data is essential for several urban understanding tasks such as reconstruction, navigation, and mapping. Despite significant progress in the area, extracting accurate polygonal building footprints remains an open problem. In this paper, we introduce Pix2Poly, an attention-based end-to-end trainable and differentiable deep neural network capable of directly generating explicit high-quality building footprints in a ring graph format. Pix2Poly employs a generative encoder-decoder transformer to produce a sequence of graph vertex tokens whose connectivity information is learned by an optimal matching network. Compared to previous graph learning methods, ours is a truly end-to-end trainable approach that extracts high-quality building footprints and road networks without requiring complicated, computationally intensive raster loss functions and intricate training pipelines. Upon evaluating Pix2Poly on several complex and challenging datasets, we report that Pix2Poly outperforms state-of-the-art methods in several vector shape quality metrics while being an entirely explicit method. Our code is available at https://github.com/yeshwanth95/Pix2Poly.

Abstract We present Im2SurfTex, a method that generates textures for input 3D shapes by learning to aggregate multi‐view image outputs produced by 2D image diffusion models onto the shapes' texture space. Unlike existing texture generation techniques that use ad hoc backprojection and averaging schemes to blend multiview images into textures, often resulting in texture seams and artifacts, our approach employs a trained neural module to boost texture coherency. The key ingredient of our module is to leverage neural attention and appropriate positional encodings of image pixels based on their corresponding 3D point positions, normals, and surface‐aware coordinates as encoded in geodesic distances within surface patches. These encodings capture texture correlations between neighboring surface points, ensuring better texture continuity. Experimental results show that our module improves texture quality, achieving superior performance in high‐resolution texture generation.

We introduce FacadeNet, a deep learning approach for synthesizing building facade images from diverse viewpoints. Our method employs a conditional GAN, taking a single view of a facade along with the desired viewpoint information and generates an image of the facade from the distinct viewpoint. To precisely modify view-dependent elements like windows and doors while preserving the structure of view-independent components such as walls, we introduce a selective editing module. This module leverages image embeddings extracted from a pretrained vision transformer. Our experiments demonstrated state-of-the-art performance on building facade generation, surpassing alternative methods.

Abstract We present a deep learning method that propagates point‐wise feature representations across shapes within a collection for the purpose of 3D shape segmentation. We propose a cross‐shape attention mechanism to enable interactions between a shape's point‐wise features and those of other shapes. The mechanism assesses both the degree of interaction between points and also mediates feature propagation across shapes, improving the accuracy and consistency of the resulting point‐wise feature representations for shape segmentation. Our method also proposes a shape retrieval measure to select suitable shapes for cross‐shape attention operations for each test shape. Our experiments demonstrate that our approach yields state‐of‐the‐art results in the popular PartNet dataset.

This paper proposes a method for automatic generation of textures for 3D city meshes in immersive urban environments. Many recent pipelines capture or synthesize large quantities of city geometry using scanners or procedural modeling pipelines. Such geometry is intricate and realistic, however the generation of photo-realistic textures for such large scenes remains a problem. We propose to generate textures for input target 3D meshes driven by the textural style present in readily available datasets of panoramic photos capturing urban environments. Re-targeting such 2D datasets to 3D geometry is challenging because the underlying shape, size, and layout of the urban structures in the photos do not correspond to the ones in the target meshes. Photos also often have objects (e.g., trees, vehicles) that may not even be present in the target geometry. To address these issues we present a method, called Projective Urban Texturing (PUT), which re-targets textural style from real-world panoramic images to unseen urban meshes. PUT relies on contrastive and adversarial training of a neural architecture designed for unpaired image-to-texture translation. The generated textures are stored in a texture atlas applied to the target 3D mesh geometry. To promote texture consistency, PUT employs an iterative procedure in which texture synthesis is conditioned on previously generated, adjacent textures. We demonstrate both quantitative and qualitative evaluation of the generated textures.

We introduce BuildingNet: (a) a large-scale dataset of 3D building models whose exteriors are consistently labeled, and (b) a graph neural network that labels building meshes by analyzing spatial and structural relations of their geometric primitives. To create our dataset, we used crowdsourcing combined with expert guidance, resulting in 513K annotated mesh primitives, grouped into 292K semantic part components across 2K building models. The dataset covers several building categories, such as houses, churches, skyscrapers, town halls, libraries, and castles. We include a benchmark for evaluating mesh and point cloud labeling. Buildings have more challenging structural complexity compared to objects in existing benchmarks (e.g., ShapeNet, PartNet), thus, we hope that our dataset can nurture the development of algorithms that are able to cope with such large-scale geometric data for both vision and graphics tasks e.g., 3D semantic segmentation, part-based generative models, correspondences, texturing, and analysis of point cloud data acquired from real-world buildings. Finally, we show that our mesh-based graph neural network significantly improves performance over several baselines for labeling 3D meshes. Our project page www.buildingnet.org includes our dataset and code.

Abstract We present a method that detects boundaries of parts in 3D shapes represented as point clouds. Our method is based on a graph convolutional network architecture that outputs a probability for a point to lie in an area that separates two or more parts in a 3D shape. Our boundary detector is quite generic: it can be trained to localize boundaries of semantic parts or geometric primitives commonly used in 3D modeling. Our experiments demonstrate that our method can extract more accurate boundaries that are closer to ground‐truth ones compared to alternatives. We also demonstrate an application of our network to fine‐grained semantic shape segmentation, where we also show improvements in terms of part labeling performance.

Material understanding is critical for design, geometric modeling, and analysis of functional objects. We enable material-aware 3D shape analysis by employing a projective convolutional neural network architecture to learn material-aware descriptors from view-based representations of 3D points for point-wise material classification or material-aware retrieval. Unfortunately, only a small fraction of shapes in 3D repositories are labeled with physical materials, posing a challenge for learning methods. To address this challenge, we crowdsource a dataset of 3080 3D shapes with part-wise material labels. We focus on furniture models which exhibit interesting structure and material variability. In addition, we also contribute a high-quality expert-labeled benchmark of 115 shapes from Herman-Miller and IKEA for evaluation. We further apply a mesh-aware conditional random field, which incorporates rotational and reflective symmetries, to smooth our local material predictions across neighboring surface patches. We demonstrate the effectiveness of our learned descriptors for automatic texturing, material-aware retrieval, and physical simulation.

This paper introduces a deep architecture for segmenting 3D objects into their labeled semantic parts. Our architecture combines image-based Fully Convolutional Networks (FCNs) and surface-based Conditional Random Fields (CRFs) to yield coherent segmentations of 3D shapes. The image-based FCNs are used for efficient view-based reasoning about 3D object parts. Through a special projection layer, FCN outputs are effectively aggregated across multiple views and scales, then are projected onto the 3D object surfaces. Finally, a surface-based CRF combines the projected outputs with geometric consistency cues to yield coherent segmentations. The whole architecture (multi-view FCNs and CRF) is trained end-to-end. Our approach significantly outperforms the existing state-of-the-art methods in the currently largest segmentation benchmark (ShapeNet). Finally, we demonstrate promising segmentation results on noisy 3D shapes acquired from consumer-grade depth cameras.

We introduce a method for co-locating style-defining elements over a set of 3D shapes. Our goal is to translate high-level style descriptions, such as “Ming” or “European” for furniture models, into explicit and localized regions over the geometric models that characterize each style. For each style, the set of style-defining elements is defined as the union of all the elements that are able to discriminate the style. Another property of the style-defining elements is that they are frequently occurring, reflecting shape characteristics that appear across multiple shapes of the same style. Given an input set of 3D shapes spanning multiple categories and styles, where the shapes are grouped according to their style labels, we perform a cross-category co-analysis of the shape set to learn and spatially locate a set of defining elements for each style. This is accomplished by first sampling a large number of candidate geometric elements and then iteratively applying feature selection to the candidates, to extract style-discriminating elements until no additional elements can be found. Thus, for each style label, we obtain sets of discriminative elements that together form the superset of defining elements for the style. We demonstrate that the co-location of style-defining elements allows us to solve problems such as style classification, and enables a variety of applications such as style-revealing view selection, style-aware sampling, and style-driven modeling for 3D shapes.

Abstract Co‐aligning a collection of shapes to a consistent pose is a common problem in shape analysis with applications in shape matching, retrieval and visualization. We observe that resolving among some orientations is easier than others, for example, a common mistake for bicycles is to align front‐to‐back, while even the simplest algorithm would not erroneously pick orthogonal alignment. The key idea of our work is to analyse rotational autocorrelations of shapes to facilitate shape co‐alignment. In particular, we use such an autocorrelation measure of individual shapes to decide which shape pairs might have well‐matching orientations; and, if so, which configurations are likely to produce better alignments. This significantly prunes the number of alignments to be examined, and leads to an efficient, scalable algorithm that performs comparably to state‐of‐the‐art techniques on benchmark data sets, but requires significantly fewer computations, resulting in 2–16× speed improvement in our tests.

Abstract Recent advances in modeling tools enable non‐expert users to synthesize novel shapes by assembling parts extracted from model databases. A major challenge for these tools is to provide users with relevant parts, which is especially difficult for large repositories with significant geometric variations. In this paper we analyze unorganized collections of 3D models to facilitate explorative shape synthesis by providing high‐level feedback of possible synthesizable shapes. By jointly analyzing arrangements and shapes of parts across models, we hierarchically embed the models into low‐dimensional spaces. The user can then use the parameterization to explore the existing models by clicking in different areas or by selecting groups to zoom on specific shape clusters. More importantly, any point in the embedded space can be lifted to an arrangement of parts to provide an abstracted view of possible shape variations. The abstraction can further be realized by appropriately deforming parts from neighboring models to produce synthesized geometry. Our experiments show that users can rapidly generate plausible and diverse shapes using our system, which also performs favorably with respect to previous modeling tools.

Abstract Extracting semantically related parts across models remains challenging, especially without supervision. The common approach is to co‐analyze a model collection, while assuming the existence of descriptive geometric features that can directly identify related parts. In the presence of large shape variations, common geometric features, however, are no longer sufficiently descriptive. In this paper, we explore an indirect top‐down approach, where instead of part geometry, part arrangements extracted from each model are compared. The key observation is that while a direct comparison of part geometry can be ambiguous, part arrangements, being higher level structures, remain consistent, and hence can be used to discover latent commonalities among semantically related shapes. We show that our indirect analysis leads to the detection of recurring arrangements of parts, which are otherwise difficult to discover in a direct unsupervised setting. We evaluate our algorithm on ground truth datasets and report advantages over geometric similarity‐based bottom‐up co‐segmentation algorithms.

We introduce a meta-representation that represents the essence of a family of shapes. The meta-representation learns the configurations of shape parts that are common across the family, and encapsulates this knowledge with a system of geometric distributions that encode relative arrangements of parts. Thus, instead of predefined priors, what characterizes a shape family is directly learned from the set of input shapes. The meta-representation is constructed from a set of co-segmented shapes with known correspondence. It can then be used in several applications where we seek to preserve the identity of the shapes as members of the family. We demonstrate applications of the meta-representation in exploration of shape repositories, where interesting shape configurations can be examined in the set; guided editing, where models can be edited while maintaining their familial traits; and coupled editing, where several shapes can be collectively deformed by directly manipulating the distributions in the meta-representation. We evaluate the efficacy of the proposed representation on a variety of shape collections.