Cyclo-ETR-CBM: VLM-VQA-Guided Concept Bottleneck Learning for Interpretable Tropical Cyclone Intensity Estimation

Published: 12 July 2026| Version 1 | DOI: 10.17632/5dgxvr68sz.1
Contributors:
Abrar Faiaz, Mirza Rajit Raihan, Md. Mafizur Rahman, Md. Bazlur Rashid, S.M. Quamrul Hasan

Description

This repository contains the data files prepared for the manuscript: Cyclo-ETR-CBM: VLM-VQA-Guided Concept Bottleneck Learning for Interpretable Tropical Cyclone Intensity Estimation The files are provided to support transparency and reproducibility of the reported experiments. The dataset includes the curated cyclone image and label files used for model development, along with the VLM-derived semantic feature files generated for the storm-wise train, validation, and test partitions.

Files

Categories

Interpretable Machine Learning

Licence