Datasets:

Kullervo
/

BRIGHT

Tasks:

Image Segmentation

Feature Extraction

Zero-Shot Classification

Modalities:

Formats:

Languages:

Size:

ArXiv:

Tags:

earth-observation

disaster-response

artificial-intelligence

building-damage-mapping

Libraries:

License:

Dataset card Data Studio Files Files and versions

Kullervo commited on May 2

Commit

f37b525

·

verified ·

1 Parent(s): 3f4621e

Update README.md

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -13,8 +13,8 @@ size_categories:
 **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 12 regions worldwide, with a particular focus on developing countries.
-* About 4,500 paired optical and SAR images containing over 350,000 building instances in BRIGHT, with a spatial resolution between 0.3 and 1 meters, provides detailed representations of individual buildings.
 **IEEE GRSS Data Fusion Contest 2025**

 **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 individual buildings.
 **IEEE GRSS Data Fusion Contest 2025**