hotosm
Datasets
All datasets matching “hotosm”venezuela_eq_2026
2026 Venezuela earthquake: AI building and damage assessment
Download the data from HDX: Venezuela - M 7.5 Earthquake - Damage Assessment Area (Validated)
AI-derived building footprints and building-level damage for the 24 June 2026 Venezuela earthquake,
organized by area. Damage comes from up to three independent AI sources: HOTOSM fAIr (primary),
Microsoft AI for Good Lab, and an OSU/CUNY Sentinel-1 radar product.
Interactive map: view it in map.
Building Damage… See the full description on the dataset page: https://huggingface.co/datasets/hotosm/venezuela_eq_2026.vhr-building-segmentation
HOT Building Segmentation Dataset
Dataset Description
A semantic segmentation dataset for building footprint extraction from aerial imagery, built from validated Humanitarian OpenStreetMap Team (HOT) Tasking Manager projects that use OpenAerialMap (OAM) imagery.
Dataset Summary
This dataset pairs 256x256 aerial image tiles (zoom level 19) from OpenAerialMap with building footprint labels from OpenStreetMap. All source projects have been fully… See the full description on the dataset page: https://huggingface.co/datasets/hotosm/vhr-building-segmentation.colombia_eq_2026nepal_flood_2026
Nepal Flood 2026, Upper Trishuli and Bhote Koshi Corridor
Building inventory for the 26 August 2026 flash flood on the Nepal-China border.
AOI: 1 km buffer around the Bhote Koshi and Trishuli river centrelines, 132.93 km² across Nuwakot and Rasuwa districts. Tasking Manager project 62904.
Contents
upperstream/buildings.geojson (+ .parquet): 13,663 footprints
upperstream/building_density_h3_r8.geojson (+ .parquet): buildings per H3 res 8 cell (~0.7 km²)… See the full description on the dataset page: https://huggingface.co/datasets/hotosm/nepal_flood_2026.streetlevel-poles
Street-level Poles & Towers
Object detection dataset of utility poles and towers in street-level
imagery. Curated by UC Berkeley as part of the HOT-OSM YOLO experiments.
This release packages the working set in three compatible layouts:
HuggingFace parquet (default loader): images embedded as bytes with
COCO-style boxes (x, y, w, h in pixels), class labels, and per-image
source metadata. Loadable with datasets.load_dataset(...).
COCO JSON under coco/instances_{train,val… See the full description on the dataset page: https://huggingface.co/datasets/hotosm/streetlevel-poles.vhr-highway-segmentation
VHR Highway Segmentation
Road centreline training data at very high resolution. Aerial imagery tiles from
OpenAerialMap paired with OpenStreetMap highway geometries, covering every validated
OAM road-mapping project on the HOT Tasking Manager from the last five years.
The building counterpart is
hotosm/vhr-building-segmentation.
Both index tiles by tile_id, so they can be joined for multi-task training.
Usage
from datasets import load_dataset
ds =… See the full description on the dataset page: https://huggingface.co/datasets/hotosm/vhr-highway-segmentation.