A Large Dataset of Object Scans

Abstract

We have created a dataset of more than ten thousand 3D scans of real objects. To create the dataset, we recruited 70 operators, equipped them with consumer-grade mobile 3D scanning setups, and paid them to scan objects in their environments. The operators scanned objects of their choosing, outside the laboratory and without direct supervision by computer vision professionals. The result is a large and diverse collection of object scans: from shoes, mugs, and toys to grand pianos, construction vehicles, and large outdoor sculptures. We worked with an attorney to ensure that data acquisition did not violate privacy constraints. The acquired data was irrevocably placed in the public domain and is available freely at http://redwood-data.org/3dscan .

Technical report

Cited by in corpus (37)

• VisDA: The Visual Domain Adaptation Challenge
• ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes
• Matterport3D: Learning from RGB-D Data in Indoor Environments
• Robust Spatial Filtering with Graph Convolutional Neural Networks
• A Rotation-Invariant Framework for Deep Point Cloud Analysis
• Single-Image Depth Perception in the Wild
• SpaceNet MVOI: a Multi-View Overhead Imagery Dataset
• Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesis
• Improved Adversarial Systems for 3D Object Generation and Reconstruction
• Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling
• Syn2Real: A New Benchmark forSynthetic-to-Real Visual Domain Adaptation
• Depth-supervised NeRF: Fewer Views and Faster Training for Free
• OctNet: Learning Deep 3D Representations at High Resolutions
• OctNetFusion: Learning Depth Fusion from Data
• DIST: Rendering Deep Implicit Signed Distance Function with Differentiable Sphere Tracing
• UnrealCV: Connecting Computer Vision to Unreal Engine
• 3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction
• FroDO: From Detections to 3D Objects
• Road Network and Travel Time Extraction from Multiple Look Angles with SpaceNet Data
• RealMonoDepth: Self-Supervised Monocular Depth Estimation for General Scenes
• From Planes to Corners: Multi-Purpose Primitive Detection in Unorganized 3D Point Clouds
• Learning to Communicate and Correct Pose Errors
• Few-Shot Point Cloud Region Annotation with Human in the Loop
• A Large-Scale Evaluation of Shape-Aware Neighborhood Weights and Neighborhood Sizes
• 3D Spatial Recognition without Spatially Labeled 3D
• Learning to Generate Synthetic 3D Training Data through Hybrid Gradient
• Learning Transformation Synchronization
• Object-based Illumination Estimation with Rendering-aware Neural Networks
• Interactive Annotation of 3D Object Geometry using 2D Scribbles
• Shape from Shading through Shape Evolution
• Self-supervised Geometric Perception
• Geometric Affordances from a Single Example via the Interaction Tensor
• Extending DeepSDF for automatic 3D shape retrieval and similarity transform estimation
• 3D Surface Reconstruction by Pointillism
• Rotation Averaging in a Split Second: A Primal-Dual Method and a Closed-Form for Cycle Graphs
• Consistent Two-Flow Network for Tele-Registration of Point Clouds
• Generic Primitive Detection in Point Clouds Using Novel Minimal Quadric Fits