Lightweight Multimodal Adaptation of Vision Language Models for Species Recognition and Habitat Context Interpretation in Drone Thermal Imagery

TL;DR

This paper introduces a lightweight multimodal adaptation framework that effectively transfers RGB-pretrained vision language models to thermal drone imagery, enabling species recognition and habitat context interpretation.

Contribution

The study develops a novel multimodal projector alignment method for adapting vision language models to thermal imagery, demonstrating improved ecological monitoring capabilities.

Findings

Qwen3-VL-8B-Instruct with open-set prompting achieved F1 scores above 0.9 for species recognition.

The models accurately enumerate instances with within-1 accuracy up to 1.0.

Combining thermal and RGB imagery enables habitat and landscape feature interpretation.

Abstract

This study proposes a lightweight multimodal adaptation framework to bridge the representation gap between RGB-pretrained VLMs and thermal infrared imagery, and demonstrates its practical utility using a real drone-collected dataset. A thermal dataset was developed from drone-collected imagery and was used to fine-tune VLMs through multimodal projector alignment, enabling the transfer of information from RGB-based visual representations to thermal radiometric inputs. Three representative models, including InternVL3-8B-Instruct, Qwen2.5-VL-7B-Instruct, and Qwen3-VL-8B-Instruct, were benchmarked under both closed-set and open-set prompting conditions for species recognition and instance enumeration. Among the tested models, Qwen3-VL-8B-Instruct with open-set prompting achieved the best overall performance, with F1 scores of 0.935 for deer, 0.915 for rhino, and 0.968 for elephant, and within-1 enumeration accuracies of 0.779, 0.982, and 1.000, respectively. In addition, combining thermal imagery with simultaneously collected RGB imagery enabled the model to generate habitat-context information, including land-cover characteristics, key landscape features, and visible human disturbance. Overall, the findings demonstrate that lightweight projector-based adaptation provides an effective and practical route for transferring RGB-pretrained VLMs to thermal drone imagery, expanding their utility from object-level recognition to habitat-context interpretation in ecological monitoring.