datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
streetview-global
StreetView Global
A globally-sampled street-view image dataset with rich scene annotations and
visual question-answer pairs. All photographs are sourced from
Mapillary, the open street-level imagery
platform, via its public image API.
Each example pairs a street-level photograph with geographic metadata
(latitude, longitude, compass, capture time, region), a free-form scene
description, structured scene classification (setting, weather, time of day,
road type, infrastructure), and… See the full description on the dataset page: https://huggingface.co/datasets/Reubencf/streetview-global.global-dense-instances
Global Dense Satellite — Instance Annotations (v1.1)
932,427 polygon-level object annotations across 275 hierarchical sites,
anchored to global-dense-satellite satellite tiles at zoom levels 10-18. Each instance is a
polygon (building footprint or OSM area feature) with structured OSM tags, projected into the
pixel frame of its deepest containing tile.
This is an annotations-only release — the satellite imagery is in
dcher95/global-dense-satellite and must be downloaded… See the full description on the dataset page: https://huggingface.co/datasets/dcher95/global-dense-instances.PPE-Global
PPE-Global
Version 1.1 — changes from v1.0 (GIFT woodiness re-fetched against every species; observer-disjoint split_observer added; every other table identical) are itemised in PROVENANCE.md §6–7.
A global plant-phenology dataset built from iNaturalist research-grade observations.
Metadata only — imagery is referenced by photo_id and fetched from the public
inaturalist-open-data S3 bucket (see Images below).
Extends its predecessor Pheno3M/PPE (3,573,640 observations, 6,825… See the full description on the dataset page: https://huggingface.co/datasets/dcher95/PPE-Global.GlobalRG-Retrieval
GlobalRG - Retrieval Across Universals Task
Despite recent advancements in vision-language models, their performance remains suboptimal on images from non-western cultures due to underrepresentation in training datasets. Various benchmarks have been proposed to test models' cultural inclusivity, but they have limited coverage of cultures and do not adequately assess cultural diversity across universal as well as culture-specific local concepts. We introduce the GlobalRG-Retrieval… See the full description on the dataset page: https://huggingface.co/datasets/UBC-VL/GlobalRG-Retrieval.srtm-3-arc-second-global
SRTM 3 Arc-Second Global
Raw ASCII heightmaps of the Earth's surface labelled according to latitude and longitude.
Mission Description
The Shuttle Radar Topography Mission (SRTM) was flown aboard the space shuttle Endeavour February 11-22, 2000. The National Aeronautics and Space Administration (NASA) and the National Geospatial-Intelligence Agency (NGA) participated in an international project to acquire radar data which were used to create the first near-global set of… See the full description on the dataset page: https://huggingface.co/datasets/mpatrick1991/srtm-3-arc-second-global.