The work creates the first dataset and baseline for generating emission textures on 3D objects to reproduce glowing materials from input images.
hub Mixed citations
ShapeNet: An Information-Rich 3D Model Repository
Mixed citation behavior. Most common role is background (55%).
93
Pith papers citing it
Background
55% of classified citations
abstract
We present ShapeNet: a richly-annotated, large-scale repository of shapes represented by 3D CAD models of objects. ShapeNet contains 3D models from a multitude of semantic categories and organizes them under the WordNet taxonomy. It is a collection of datasets providing many semantic annotations for each 3D model such as consistent rigid alignments, parts and bilateral symmetry planes, physical sizes, keywords, as well as other planned annotations. Annotations are made available through a public web-based interface to enable data visualization of object attributes, promote data-driven geometric analysis, and provide a large-scale quantitative benchmark for research in computer graphics and vision. At the time of this technical report, ShapeNet has indexed more than 3,000,000 models, 220,000 models out of which are classified into 3,135 categories (WordNet synsets). In this report we describe the ShapeNet effort as a whole, provide details for all currently available datasets, and summarize future plans.
hub tools
citation-role summary
dataset 12
background 7
method 1
citation-polarity summary
claims ledger
- abstract We present ShapeNet: a richly-annotated, large-scale repository of shapes represented by 3D CAD models of objects. ShapeNet contains 3D models from a multitude of semantic categories and organizes them under the WordNet taxonomy. It is a collection of datasets providing many semantic annotations for each 3D model such as consistent rigid alignments, parts and bilateral symmetry planes, physical sizes, keywords, as well as other planned annotations. Annotations are made available through a public web-based interface to enable data visualization of object attributes, promote data-driven geometri
co-cited works
representative citing papers
ARKitScenes is the largest real-world indoor RGB-D dataset captured with mobile LiDAR, including high-resolution depth maps and 3D furniture bounding box annotations for advancing object detection and depth upsampling.
MAPS provides 2618 validated 3D meshes and a controllable rendering pipeline to attribute vision model recognition failures to specific scene parameters, finding camera distance and elevation as the dominant failure factors across 20 tested models.
OffsetAxis reconstructs meshes from unsigned distance fields by extracting the medial axis of the alpha-offset volume using ray casting and variational medial ball optimization.
min-GSGW learns coupled nonlinear slicers to produce a rigid-motion-invariant, scalable approximation to the Gromov-Wasserstein distance and its transport plans.
Img2CADSeq generates standard CAD sequences from images via a multi-stage pipeline with three-level hierarchical codebook encoding, importance-guided compression, and contrastive point-cloud conditioning of a VQ-Diffusion model, outperforming prior methods on new CAD-220K and PrintCAD datasets.
Multi-grained counting is introduced with five granularity levels, supported by the new KubriCount dataset generated via 3D synthesis and editing, and HieraCount model that combines text and visual exemplars for improved accuracy.
Language representations serve as the asymptotic attractor for convergence in independently trained multimodal neural networks due to feature density asymmetry.
MeshFIM enables local low-poly mesh editing by autoregressively filling target regions conditioned on context, using boundary markers, positional embeddings, and a gated geometry encoder to enforce attachment, topology, and region limits.
Reinforcement learning internalizes physical stability rules for brick structures, enabling the first rollback-free generation with orders-of-magnitude faster inference.
Consistency learning reformulates 3D point cloud anomaly detection to predict clean geometry directly in one or two steps, yielding up to 80 times faster inference while matching state-of-the-art accuracy.
ADS adaptively refines a Delaunay scaffold to produce unbiased random samples on occupancy function surfaces together with a connecting mesh, using far fewer evaluations than existing approaches.
OGPP is a particle flow-matching method using orbit-space canonicalization and geometric paths that achieves lower error and fewer steps than prior approaches on 3D benchmarks.
AirZoo is a new large-scale synthetic dataset for aerial 3D vision that improves state-of-the-art models on image retrieval, cross-view matching, and 3D reconstruction when used for fine-tuning.
Topo-ADV uses differentiable persistent homology to create topology-altering perturbations that achieve up to 100% attack success on point cloud classifiers like PointNet while remaining geometrically imperceptible.
XShapeEnc encodes arbitrary 2D spatially grounded shapes into compact invertible representations by decomposing them into unit-disk geometry and harmonic pose fields then applying Zernike bases with frequency propagation.
3D-Fixer performs in-place 3D asset completion from single-view partial point clouds via coarse-to-fine generation with ORFA conditioning, plus a new ARSG-110K dataset, to achieve higher geometric accuracy than MIDI and Gen3DSR while keeping diffusion efficiency.
DeformPIC deforms query point clouds under prompt guidance for in-context learning, outperforming prior methods with lower Chamfer Distance on reconstruction, denoising, and registration tasks.
PointATA is a parameter-efficient transfer learning method that aligns 3D-4D modality gaps via optimal transport before adapting a frozen 3D model with video-specific modules to achieve strong 4D perception results.
CLIPoint3D is the first CLIP-based framework for few-shot unsupervised 3D point cloud domain adaptation that reports 3-16% accuracy gains on PointDA-10 and GraspNetPC-10.
A method estimates mass from single RGB images by fusing depth-based volume cues with vision-language model density semantics via adaptive gating and separate regression heads trained on mass labels only.
A low-memory streaming estimator for sliced Wasserstein distance using quantile approximations on random projections with theoretical error guarantees.
A spectrum-aware decision boundary algorithm enables effective hard-label black-box adversarial attacks on 3D point cloud models by fusing spectral information across classes and performing curvature-aware iterative optimization.
LRM is a large transformer that predicts a NeRF directly from a single image after training on a million-object multi-view dataset.
citing papers explorer
-
Towards Realistic 3D Emission Materials: Dataset, Baseline, and Evaluation for Emission Texture Generation
The work creates the first dataset and baseline for generating emission textures on 3D objects to reproduce glowing materials from input images.
-
ARKitScenes: A Diverse Real-World Dataset For 3D Indoor Scene Understanding Using Mobile RGB-D Data
ARKitScenes is the largest real-world indoor RGB-D dataset captured with mobile LiDAR, including high-resolution depth maps and 3D furniture bounding box annotations for advancing object detection and depth upsampling.
-
MAPS: A Synthetic Dataset for Probing Vision Models in a Controlled 3D Scene Space
MAPS provides 2618 validated 3D meshes and a controllable rendering pipeline to attribute vision model recognition failures to specific scene parameters, finding camera distance and elevation as the dominant failure factors across 20 tested models.
-
OffsetAxis: UDF Mesh Reconstruction via Offset-Volume Medial Axis Extraction
OffsetAxis reconstructs meshes from unsigned distance fields by extracting the medial axis of the alpha-offset volume using ray casting and variational medial ball optimization.
-
Min Generalized Sliced Gromov Wasserstein: A Scalable Path to Gromov Wasserstein
min-GSGW learns coupled nonlinear slicers to produce a rigid-motion-invariant, scalable approximation to the Gromov-Wasserstein distance and its transport plans.
-
Img2CADSeq: Image-to-CAD Generation via Sequence-Based Diffusion
Img2CADSeq generates standard CAD sequences from images via a multi-stage pipeline with three-level hierarchical codebook encoding, importance-guided compression, and contrastive point-cloud conditioning of a VQ-Diffusion model, outperforming prior methods on new CAD-220K and PrintCAD datasets.
-
Count Anything at Any Granularity
Multi-grained counting is introduced with five granularity levels, supported by the new KubriCount dataset generated via 3D synthesis and editing, and HieraCount model that combines text and visual exemplars for improved accuracy.
-
The Wittgensteinian Representation Hypothesis: Is Language the Attractor of Multimodal Convergence?
Language representations serve as the asymptotic attractor for convergence in independently trained multimodal neural networks due to feature density asymmetry.
-
MeshFIM: Local Low-Poly Mesh Editing via Fill-in-the-Middle Autoregressive Generation
MeshFIM enables local low-poly mesh editing by autoregressively filling target regions conditioned on context, using boundary markers, positional embeddings, and a gated geometry encoder to enforce attachment, topology, and region limits.
-
Rollback-Free Stable Brick Structures Generation
Reinforcement learning internalizes physical stability rules for brick structures, enabling the first rollback-free generation with orders-of-magnitude faster inference.
-
Two Steps Are All You Need: Efficient 3D Point Cloud Anomaly Detection with Consistency Models
Consistency learning reformulates 3D point cloud anomaly detection to predict clean geometry directly in one or two steps, yielding up to 80 times faster inference while matching state-of-the-art accuracy.
-
ADS: Random Sampling of Occupancy Functions using Adaptive Delaunay Scaffolding
ADS adaptively refines a Delaunay scaffold to produce unbiased random samples on occupancy function surfaces together with a connecting mesh, using far fewer evaluations than existing approaches.
-
Generative Modeling with Orbit-Space Particle Flow Matching
OGPP is a particle flow-matching method using orbit-space canonicalization and geometric paths that achieves lower error and fewer steps than prior approaches on 3D benchmarks.
-
AirZoo: A Unified Large-Scale Dataset for Grounding Aerial Geometric 3D Vision
AirZoo is a new large-scale synthetic dataset for aerial 3D vision that improves state-of-the-art models on image retrieval, cross-view matching, and 3D reconstruction when used for fine-tuning.
-
Topo-ADV: Generating Topology-Driven Imperceptible Adversarial Point Clouds
Topo-ADV uses differentiable persistent homology to create topology-altering perturbations that achieve up to 100% attack success on point cloud classifiers like PointNet while remaining geometrically imperceptible.
-
Training-free Spatially Grounded Geometric Shape Encoding (Technical Report)
XShapeEnc encodes arbitrary 2D spatially grounded shapes into compact invertible representations by decomposing them into unit-disk geometry and harmonic pose fields then applying Zernike bases with frequency propagation.
-
3D-Fixer: Coarse-to-Fine In-place Completion for 3D Scenes from a Single Image
3D-Fixer performs in-place 3D asset completion from single-view partial point clouds via coarse-to-fine generation with ORFA conditioning, plus a new ARSG-110K dataset, to achieve higher geometric accuracy than MIDI and Gen3DSR while keeping diffusion efficiency.
-
Deformation-based In-Context Learning for Point Cloud Understanding
DeformPIC deforms query point clouds under prompt guidance for in-context learning, outperforming prior methods with lower Chamfer Distance on reconstruction, denoising, and registration tasks.
-
Align then Adapt: Rethinking Parameter-Efficient Transfer Learning in 4D Perception
PointATA is a parameter-efficient transfer learning method that aligns 3D-4D modality gaps via optimal transport before adapting a frozen 3D model with video-specific modules to achieve strong 4D perception results.
-
CLIPoint3D: Language-Grounded Few-Shot Unsupervised 3D Point Cloud Domain Adaptation
CLIPoint3D is the first CLIP-based framework for few-shot unsupervised 3D point cloud domain adaptation that reports 3-16% accuracy gains on PointDA-10 and GraspNetPC-10.
-
Physically Guided Visual Mass Estimation from a Single RGB Image
A method estimates mass from single RGB images by fusing depth-based volume cues with vision-language model density semantics via adaptive gating and separate regression heads trained on mass labels only.
-
Streaming Sliced Optimal Transport
A low-memory streaming estimator for sliced Wasserstein distance using quantile approximations on random projections with theoretical error guarantees.
-
Hard-Label Black-Box Attacks on 3D Point Clouds
A spectrum-aware decision boundary algorithm enables effective hard-label black-box adversarial attacks on 3D point cloud models by fusing spectral information across classes and performing curvature-aware iterative optimization.
-
LRM: Large Reconstruction Model for Single Image to 3D
LRM is a large transformer that predicts a NeRF directly from a single image after training on a million-object multi-view dataset.
-
Objaverse-XL: A Universe of 10M+ 3D Objects
Objaverse-XL supplies over 10 million diverse 3D objects that, when used to render 100 million views, improve zero-shot novel-view synthesis in models such as Zero123.
-
Fast Graph Representation Learning with PyTorch Geometric
PyTorch Geometric is a PyTorch library that delivers fast graph neural network training through sparse GPU kernels and variable-size mini-batching.
-
Rethinking Point Clouds as Sequences: A Causal Next-Token Predictive Learning Framework
PointNTP serializes point clouds into geometry-ordered patch sequences and applies causal next-token prediction with stop-gradient targets for decoder-free self-supervised pre-training, reporting competitive results on ScanObjectNN, ShapeNetPart, and S3DIS.
-
A Holistic Method for Superquadric Fitting Using Unsupervised Clustering Analysis
Introduces a clustering-based optimization technique for fitting superquadrics to point clouds that handles noise, outliers, and deformations with closed-form solutions and convergence proofs.
-
Sat3DGen: Comprehensive Street-Level 3D Scene Generation from Single Satellite Image
Sat3DGen improves geometric RMSE from 6.76m to 5.20m and FID from ~40 to 19 for street-level 3D generation from satellite images via geometry-centric constraints and perspective training.
-
ObjView-Bench: Rethinking Difficulty and Deployment for Object-Centric View Planning
ObjView-Bench disentangles omnidirectional self-occlusion, saturation difficulty, and set-cover planning difficulty, then shows that budget regimes and reachable-view constraints change planner rankings and failure modes across classical, learned, and hybrid methods.
-
GenMed: A Pairwise Generative Reformulation of Medical Diagnostic Tasks
GenMed uses diffusion models to capture P(X,Y) for medical tasks and performs inference via gradient-based test-time optimization, supporting arbitrary observation combinations without retraining.
-
Beyond Spatial Compression: Interface-Centric Generative States for Open-World 3D Structure
C2LT-3D factorizes 3D tokenization into canonical local geometry, partition-conditioned context, and relational seam variables to make latent states operational for assembly-level validation and repair in open-world multi-component assets.
-
Minimax Optimal Estimation of Transport-Growth Pairs in Unbalanced Optimal Transport
Estimators for transport-growth pairs in unbalanced OT achieve minimax optimal rates, supported by a value-based stability reduction through a UOT gap condition.
-
Toward Visually Realistic Simulation: A Benchmark for Evaluating Robot Manipulation in Simulation
VISER is a new visually realistic simulation benchmark for robot manipulation tasks that uses PBR materials and MLLM-assisted asset generation, achieving 0.92 Pearson correlation with real-world policy performance.
-
Prop-Chromeleon: Adaptive Haptic Props in Mixed Reality through Generative Artificial Intelligence
A generative-AI pipeline dynamically generates and anchors virtual assets to match the shape of physical props, enabling adaptive passive haptics in MR that users rate higher in realism, immersion, and enjoyment than static baselines.
-
TAFA-GSGC: Group-wise Scalable Point Cloud Geometry Compression with Progressive Residual Refinement
TAFA-GSGC is a scalable point cloud geometry compression codec using progressive residual refinement and group-wise entropy coding that achieves average BD-rate reductions of 4.99% (D1-PSNR) and 5.92% (D2-PSNR) over PCGCv2 while supporting monotonic multi-quality decoding from a single bitstream.
-
ShapeY: A Principled Framework for Measuring Shape Recognition Capacity via Nearest-Neighbor Matching
ShapeY is a benchmark dataset and nearest-neighbor protocol that measures shape-based recognition in vision models, revealing that even state-of-the-art networks fail to generalize consistently across 3D viewpoints and non-shape appearance changes.
-
Point-MF: One-step Point Cloud Generation from a Single Image via Mean Flows
Point-MF performs one-step point cloud reconstruction from single images by learning a mean velocity field in point space with a tailored Diffusion Transformer and a new auxiliary loss.
-
Text-Guided Multimodal Unified Industrial Anomaly Detection
A text-semantics-guided multimodal framework with geometry-aware mapping and object-conditioned text adaptation achieves state-of-the-art unsupervised anomaly detection and localization on RGB-3D industrial datasets while enabling a single model for multiple classes.
-
FILTR: Extracting Topological Features from Pretrained 3D Models
FILTR predicts persistence diagrams from pretrained 3D encoders on the new DONUT benchmark, showing limited topological signals in encoders but successful approximation via learnable feed-forward.
-
FurnSet: Exploiting Repeats for 3D Scene Reconstruction
FurnSet improves single-view 3D scene reconstruction by using per-object CLS tokens and set-aware self-attention to group and jointly reconstruct repeated object instances, with added scene-object conditioning and layout optimization.
-
Volume Transformer: Revisiting Vanilla Transformers for 3D Scene Understanding
A minimally modified vanilla Transformer called Volt achieves state-of-the-art 3D semantic and instance segmentation by using volumetric tokens, 3D rotary embeddings, and a data-efficient training recipe that scales better than domain-specific backbones.
-
One-Shot Cross-Geometry Skill Transfer through Part Decomposition
Part decomposition with generative shape models allows one-shot robot skill transfer across unfamiliar object geometries in simulation and real settings.
-
Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective
The paper proposes a problem-driven taxonomy for feed-forward 3D scene modeling that groups methods by five core challenges: feature enhancement, geometry awareness, model efficiency, augmentation strategies, and temporal-aware modeling.
-
ReplicateAnyScene: Zero-Shot Video-to-3D Composition via Textual-Visual-Spatial Alignment
ReplicateAnyScene performs fully automated zero-shot video-to-compositional-3D reconstruction by cascading alignments of generic priors from vision foundation models across textual, visual, and spatial dimensions.
-
L-PCN: A Point Cloud Accelerator Exploiting Spatial Locality through Octree-based Islandization
L-PCN exploits spatial locality in point cloud networks via octree partitioning into islands and intra-island hub scheduling, delivering 55-94% less feature fetching, 45-81% less computation, and 1.2-3.2x additional speedup on FPGA prototypes.
-
TouchAnything: Diffusion-Guided 3D Reconstruction from Sparse Robot Touches
TouchAnything reconstructs accurate 3D object geometries from only a few tactile contacts by optimizing for consistency with a pretrained visual diffusion prior.
-
Part-Level 3D Gaussian Vehicle Generation with Joint and Hinge Axis Estimation
A new framework generates part-level animatable 3D Gaussian vehicles from images by adding modules for exclusive part ownership and kinematic joint/axis prediction.
-
FusionBERT: Multi-View Image-3D Retrieval via Cross-Attention Visual Fusion and Normal-Aware 3D Encoder
FusionBERT uses cross-attention to fuse multi-view images and a normal-aware encoder for 3D models, achieving higher image-3D retrieval accuracy than prior multimodal models in both single- and multi-view settings.
-
PhysSkin: Real-Time and Generalizable Physics-Based Animation via Self-Supervised Neural Skinning
PhysSkin uses a neural skinning autoencoder and physics-informed self-supervised training to create mesh-free, generalizable skinning fields for real-time animation.