Evaluating VLM Reasoning and Localization for Floodwater Depth Estimation with Bounding Boxes
Authors:
- 3 December 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
Vision-language models (VLMs) offer a promising approach to estimate floodwater depth from images, a critical metric for emergency response. While their accuracy has been benchmarked against human estimators, their reasoning has mostly been assessed qualitatively and post hoc. In this work, we hypothesize that bounding boxes provide a systematic framework to evaluate VLM reasoning and localization. To test this, we inferenced the GLM-4.5V VLM on a dataset of flooding images and compared its depth estimates and reference-object bounding boxes against the consensus of four human annotators under four perturbation scenarios. We employ the distance intersection over union (DIoU) metric to quantify the agreement between bounding boxes, and the mean absolute error (MAE) to characterize the accuracy of floodwater depth estimates. Results show that VLMs use bounding boxes with accuracy comparable to textual descriptions, though below human performance, and that numerical estimates depend on the visual content of the chosen boxes. This demonstrates that bounding boxes can capture VLM reasoning traces and systematically reveal failure cases previously identified only through qualitative analysis.Keywords
This publication has 5 references indexed in Scilit:
- Assessing Large Multimodal Models for Urban Floodwater Depth EstimationWater Resources Research, 2025
- Urban Road Anomaly Monitoring Using Vision–Language Models for Enhanced Safety ManagementApplied Sciences, 2025
- State-of-the-Art Techniques for Real-Time Monitoring of Urban Flooding: A ReviewWater, 2024
- Automated floodwater depth estimation using large multimodal model for rapid flood mappingComputational Urban Science, 2024
- Weighted boxes fusion: Ensembling boxes from different object detection modelsImage and Vision Computing, 2021