datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
SpatialVIDSpatialVID: A Large-Scale Video Dataset with Spatial Annotations
Jiahao Wang1*
Yufeng Yuan1*
Rujie Zheng1*
Youtian Lin1
Jian Gao1
Lin-Zhuo Chen1
Yajie Bao1
Yi Zhang1
Chang Zeng1
Yanxi Zhou1
Xiaoxiao Long1
Hao Zhu1
Zhaoxiang Zhang2
Xun Cao1
Yao Yao1†
1Nanjing University 2Institute of Automation, Chinese Academy of Science
*Equal Contribution †Corresponding Author
CVPR 2026… See the full description on the dataset page: https://huggingface.co/datasets/SpatialVID/SpatialVID.SpatialVID-HQSpatialVID: A Large-Scale Video Dataset with Spatial Annotations
Jiahao Wang1*
Yufeng Yuan1*
Rujie Zheng1*
Youtian Lin1
Jian Gao1
Lin-Zhuo Chen1
Yajie Bao1
Yi Zhang1
Chang Zeng1
Yanxi Zhou1
Xiaoxiao Long1
Hao Zhu1
Zhaoxiang Zhang2
Xun Cao1
Yao Yao1†
1Nanjing University 2Institute of Automation, Chinese Academy of Science
*Equal Contribution †Corresponding Author
CVPR 2026… See the full description on the dataset page: https://huggingface.co/datasets/FelixYuan/SpatialVID-HQ.SpatialLM-Testset
SpatialLM Testset
Project page | Paper | Code
We provide a test set of 107 preprocessed point clouds and their corresponding GT layouts, point clouds are reconstructed from RGB videos using MASt3R-SLAM. SpatialLM-Testset is quite challenging compared to prior clean RGBD scan datasets due to the noises and occlusions in the point clouds reconstructed from monocular RGB videos.
Folder Structure
Outlines of the dataset files:… See the full description on the dataset page: https://huggingface.co/datasets/manycore-research/SpatialLM-Testset.SpatialLM-Dataset
SpatialLM Dataset
The SpatialLM dataset is a large-scale, high-quality synthetic dataset designed by professional 3D designers and used for real-world production. It contains point clouds from 12,328 diverse indoor scenes comprising 54,778 rooms, each paired with rich ground-truth 3D annotations. SpatialLM dataset provides an additional valuable resource for advancing research in indoor scene understanding, 3D perception, and… See the full description on the dataset page: https://huggingface.co/datasets/manycore-research/SpatialLM-Dataset.Spatial-DISE
Spatial-DISE: A Unified Benchmark for Evaluating Spatial Reasoning in Vision-Language Models
📋 Overview
Spatial-DISE is a comprehensive benchmark dataset designed to evaluate spatial reasoning capabilities in vision-language models. The dataset focuses on various aspects of spatial intelligence including 3D perception, spatial transformation, and geometric reasoning across multiple difficulty levels.
🧪 Evaluation Support
Supported:… See the full description on the dataset page: https://huggingface.co/datasets/TACPS-liv/Spatial-DISE.spatial457_mcqSpatialConsistency-Navigationspatial457_omniView-Spatial-BenchSpatialReasoning
The Spatial Reasoning Dataset
The Spatial Reasoning Dataset comprises semantically meaningful question-answer pairs focused on the relative locations of geographic divisions within the United States — including states, counties, and ZIP codes.
The dataset is designed for spatial question answering and includes three types of questions:
Binary (Yes/No)
Single-choice (Radio)
Multi-choice (Checkbox)
All questions include correct answers for training and evaluation purposes.… See the full description on the dataset page: https://huggingface.co/datasets/Rammen/SpatialReasoning.Omni-Spatial-Benchandrade-law-saint-paul-spatial-index
Andrade Law — Saint Paul Service-Area Spatial Index
Open spatial-reference data for Andrade Law PLLC, a personal-injury law firm in Saint Paul, Minnesota. It maps the firm's office and its Saint Paul service-area landmarks to their S2 Geometry cells and WGS84 coordinates.
S2 cells are an open geometric indexing system; they are used here as geographic reference labels, not as an official or administrative identifier.
Files
Canonical home: these files are… See the full description on the dataset page: https://huggingface.co/datasets/Gabe-Andrade-Attorney/andrade-law-saint-paul-spatial-index.SpatialReasonerEvalSpatialEpiBench
SpatialEpiBench
Dataset Summary
SpatialEpiBench is a benchmark collection of 11 spatiotemporal epidemic forecasting datasets. The benchmark covers multiple public-health surveillance modalities, including influenza-like illness surveillance rates, confirmed cases, test positivity, inpatient and outpatient hospitalizations, hospital admissions, doctor visits, and deaths. The datasets span the United States, Canada, and Australia, with daily or weekly temporal resolution… See the full description on the dataset page: https://huggingface.co/datasets/ruiqil/SpatialEpiBench.Spatial_Intelligence_UnderstandingSpatialChain-Benchmark
SpatialChain Benchmark
SpatialChain is a dataset of 28,350 training and 899 test examples pairing spatially-oriented questions from GQA with scene-graph-grounded chain-of-thought reasoning traces. It enables a two-axis evaluation of thinking-enabled VLMs: standard VQA accuracy and faithfulness of the reasoning chain against the symbolic scene-graph ground truth — exposing shortcut behaviour even when the final answer is correct.
🤗 Fine-tuned model:… See the full description on the dataset page: https://huggingface.co/datasets/spatialchain/SpatialChain-Benchmark.vn-provinces-spatial-living-cost-index
Vietnam provinces spatial living cost index (Ha Noi = 100)
Spatial living-cost (sinh hoạt) price index by province relative to Ha Noi (= 100). Coverage 2011-2024. Provinces only (no regional or national aggregate in the source table). Geographic labels are English (UN/GSO style ASCII romanization). Province names follow ar_core.vn_geo (historical 63-province system).
Figures
Hero
Comparison
Color key
Files
provinces (882 rows)
data/provinces.csv… See the full description on the dataset page: https://huggingface.co/datasets/letrinhan/vn-provinces-spatial-living-cost-index.SpatialVision-dataspatial457_otchess_spatial_reasoning_10kQ-Spatial-Bench-sMAPE-Comparison
Q-Spatial-Bench-sMAPE-Comparison
Recording sMAPE for models evaluated on Q-Spatial-Bench across model families, sizes, finetuning, prompting, post-training optimization strategies.
HEST_Xenium_virtual_spatial_transcriptomics
HEST Xenium virtual spatial transcriptomics
This repository contains predicted spatial transcriptomics for HEST Xenium H&E
slides produced with DeepSpot-M.
Authors: Kalin Nonchev, Sebastian Dawo, Karina Silina, Viktor Hendrik
Koelzer, and Gunnar Rätsch.
Paper: DeepSpot-M: a multimodal foundation model for transcriptome-wide virtual spatial transcriptomics from histology (medRxiv, 2026; see the citation below).
Code: https://github.com/ratschlab/DeepSpotM.
News… See the full description on the dataset page: https://huggingface.co/datasets/ratschlab/HEST_Xenium_virtual_spatial_transcriptomics.Temporal_Spatial_Tracking_Dataset
Dataset Overview
This dataset contains time-stamped spatial tracking records collected from tagged entities (e.g., wearable tags, assets, or devices) operating within a monitored environment.Each row represents a single localization event captured at a precise moment in time, including 3D position coordinates and device status information.
The dataset is inherently temporal and spatial, making it suitable for trajectory reconstruction, movement analysis, and time-based behavioral… See the full description on the dataset page: https://huggingface.co/datasets/VillanovaAI/Temporal_Spatial_Tracking_Dataset.spatial-geometryspatial457_cleanarchitectural-spatial-blindspots-smolvlm
Architectural and Spatial Blind Spots of SmolVLM
Model Tested: HuggingFaceTB/SmolVLM-BaseParameter Count: 2.2B
1. How I loaded the model (Python Code)
I loaded the model in a Google Colab environment using the transformers library with 4-bit quantization to fit within a free-tier GPU.
from transformers import AutoProcessor, AutoModelForVision2Seq
import torch
model_id = "HuggingFaceTB/SmolVLM-Base"
processor = AutoProcessor.from_pretrained(model_id)
model =… See the full description on the dataset page: https://huggingface.co/datasets/AsefaH/architectural-spatial-blindspots-smolvlm.spatial457_testspatialagent_human
SpatialAgent — Human Expert Reference Data
Anonymized reference data produced by human scientists for two spatial-transcriptomics
tasks used to benchmark SpatialAgent:
Gene panel design — expert-designed targeted gene panels for the human
dorsolateral prefrontal cortex (DLPFC / PFC).
Cell-type & tissue-niche annotation — expert annotations of a developing human
heart MERFISH dataset (228,633 cells × 238 genes).
All scientist identities are removed. Each task uses its own… See the full description on the dataset page: https://huggingface.co/datasets/hansen7/spatialagent_human.spatial457_minispatial_geometry