Loading...
Loading...
3D models, rendered datasets, physics simulation, digital twins, synthetic data generation, game engine data
1,020 datasets
239.3 hours of first-person video data for egocentric vision and tracking. The dataset features multimodal data from egocentric views, including annotations and motion capture for extracting 3D poses, and provides detailed 3D objects and scenes using visual data from VR headsets. It was authored by UniDataPro and last updated on 2026-05-12.
Approximately 300 km offshore of San Francisco, this dataset provides high-resolution airborne LiDAR point clouds from the Sub-Mesoscale Ocean Dynamics Experiment (S-MODE). It captures ocean surface topography during a pilot campaign in October 2021 and two intensive operating periods in Fall 2022 and Spring 2023. The data aims to understand how short-scale ocean dynamics influence vertical exchanges of physical and biological variables.
Over 15,000 seabed samples provide textural composition, particle size statistics, and carbonate values for the Australian marine region. Geoscience Australia extracted this data from its Marine Sediment Database, which holds approximately 2.6 million entries. The dataset was updated in April 2026.
A dataset for training and evaluating object recognition and segmentation models on infrastructure relevant to blind and low-vision (BLV) navigation. It contains three configurations, including photorealistic IsaacSim renders for training and real photographs captured by the authors. The dataset was created by NavAble and last updated on May 7, 2026.
Jigang Li's study collected 482 valid survey responses from university basketball players. The dataset analyzes the mediated chain effect of VR technology use on training satisfaction through self-efficacy and sport engagement. It contains results from four validated scales measuring these psychological and behavioral constructs.
TerraSky3D provides multi-view reconstructions of European landmarks. The dataset was created by researchers from Graz University of Technology and Sony, and was presented at the 3DMV workshop at CVPR in 2026. It includes images collected from both aerial and ground perspectives.
RGBench Cloth Sim-to-Real (v1) contains nine carefully captured garments, each manipulated with three bimanual actions (fling, fold, grasp). The dataset provides real-world ground truth point clouds for evaluating the sim-to-real gap of cloth simulators. It was released by author hwk0809 as part of the AAAI 2026 paper Real Garment Benchmark (RGBench).
A 1500-year chronology of coastal sediment accumulation is preserved in beach deposits at Keppel Bay, Queensland, Australia. The dataset likely contains measurements of ridge morphology, sediment texture, and geochemistry, built using optically stimulated luminescence (OSL) dating. It was contributed by the Australian Ocean Data Network and last updated in April 2026.
A synthetic dataset of 10,000 images in .jpg format, each with a resolution of 64 by 64 pixels. It contains 1,000 images for each digit from 0 to 9, all presented on a black background. The dataset was created by author MishaGGG and was last updated on the platform in June 2026.
3DCode is an open-source reproduction of the 3DCodeBench academic benchmark. It contains datasets for evaluating models that generate Blender 5.0 Python code to procedurally build specific 3D objects. The benchmark includes 212 categories, each with a canonical seed, a reference factory, and two text prompts.
Multi-view multi-light HDR captures of cloth materials with associated 3D geometry, designed for inverse rendering and BRDF estimation. The dataset contains 500 materials, totaling 3.62 TB across 9,853 files, and was created by koalapenguin. It was last updated on 2026-05-03.
Montreal's Le Sud-Ouest and Ville-Marie boroughs are represented by a digital 3D building model in CityGML LOD2 format with textures. The dataset includes models in GML and 3DM formats, accompanied by a digital terrain model for the same territory. It was published by the Government and Municipalities of Québec.
PseudoKitchens is a synthetic dataset of photorealistic 3D kitchen renders with ground-truth concept annotations. It was created by author yoda-chicken and last updated on 2026-05-20. The dataset is designed for tasks such as recipe classification and spatial concept localisation.
qian43 provides supplementary files for the VIGOR dataset and the Sat3DGen paper. The supplement includes satellite depth maps, sky masks for panoramic images, and high-resolution Digital Surface Model data for Seattle. The dataset page was last updated on 2026-05-15.
Synthetic Rayleigh wave phase delays measured on vertical component seismograms for earthquake-receiver pairs across the Pacific hemisphere. The dataset includes a 3D model of vertically polarised shear wave velocity (Vsv) structure, with full 3D resolution and uncertainty, on a global grid with 2-degree lateral and 25 km vertical voxels from the surface to ~400 km depth. This data and inversion results are related to a 2025 paper in Seismica by Latallerie et al. and were last updated on 2026-04-09.
Synthetic Rayleigh wave phase delays and a 3D model of vertically polarised shear wave velocity (Vsv) for the Pacific upper mantle. The data is related to a 2025 paper by Latallerie et al. in Seismica and was last updated on the platform in April 2026. The 3D tomographic grid covers the entire sphere from the surface to ~400km depth with voxels of 2 by 2 degrees laterally and 25 km vertically.
Western Australia's Paterson Orogen is represented by a three-dimensional volumetric model constructed using 3D Geomodeller software. The model was built by the Paterson Project team and specialists from Geoscience Australia, documented in a 2026 report. It includes exported two-dimensional sections, maps, and images in Virtual Reality Modelling Language (VRML) format.
Modeled active-layer depth values in centimeters cover Russia from 50°N to 80°N and 60°E to 160°E. The dataset provides depth calculations for both current climate conditions and a projected 2050 climate scenario. It is authored by NSIDCV0 and available in tab-delimited ASCII text and GIF image formats.
A geospatial map illustrates the distribution of soil parent material textures across Alberta's agricultural region. The map classifies soil textures based on the relative proportions of sand, silt, and clay particles using a standard Soil Texture Triangle. This resource was created by the Government of Alberta in 2002 using ArcGIS.
25,000 synthetic examples in OpenAI-compatible chat format designed for supervised fine-tuning. Created by WithinUsAI in May 2026, this dataset aims to help LLMs mirror the behavior, style, tone, and capabilities of Microsoft Copilot.