Skip to content
← Directory

dataset · Global

BRIGHT: A globally distributed multimodal building damage assessment dataset with very-high-resolution for all-weather disaster response

Overview BRIGHT is the first open-access, globally distributed, event-diverse multimodal dataset specifically curated to support AI-based disaster response. It covers five types of natural disasters and two types of man-made disasters across 14 regions worldwide, with a particular focus on developing countries. About 4,200 paired optical and SAR images containing over 380,000 building instances in BRIGHT, with a spatial resolution between 0.3 and 1 meters, provides detailed representations of i

Visit source ↗

Feature or claim this listing

Feature this listing — founding Featured slots are 50% off for 6 months

Represent BRIGHT: A globally distributed multimodal building damage assessment dataset with very-high-resolution for all-weather disaster response? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Explore paid sponsorship →

Licence
cc-by-4.0
Authentication
Unknown — review source docs
GitHub stars
Unknown
Last commit
Unknown
Last successful page check
16/10/2026
Page health
restricted
Package
Not recorded
Verification
Not independently verified

Provenance & evidence

Listed in zenodo public catalogue API on 2026-10-16; relevance/value screened (haiku).

Metadata observed 14/10/2026 · research sweep: sweep-2026-10-16 (zenodo)