Benchmarking Vision-Language Models for Microscopic Plant Image Understanding

2026-06-21

AI Summary

Current vision-language models lack evaluation benchmarks tailored to microscopic plant imagery, hindering their application in plant biology and pathology at cellular and subcellular resolutions. This work introduces PlantMicro, the first benchmark specifically designed for this domain, comprising over 5,000 microscopic images spanning multiple host species, biological kingdoms, and imaging modalities, accompanied by more than 9,000 structured visual question-answer pairs. The benchmark systematically evaluates model capabilities in fine-grained recognition and biological reasoning. Experimental results reveal a significant performance gap among state-of-the-art models—e.g., GPT-5 achieves only 34.93% accuracy—highlighting both the critical need for standardized evaluation in this niche and the substantial limitations of current multimodal architectures in handling domain-specific scientific visual understanding tasks.

Abstract

Microscopic imaging provides essential visual evidence for studying plant biology and pathology at the cellular and subcellular levels. However, existing benchmarks on vision-language models primarily focus on macroscopic plant imagery, while the microscopic domain remains underexplored. To address this gap, we present PlantMicro, a comprehensive benchmark for evaluating vision-language models (VLMs) in microscopic plant imagery. PlantMicro integrates more than 5,000 images collected across diverse hosts, biological domains, and imaging modalities. Building on this diversity, we design a set of complementary tasks that capture different facets of microscopic image understanding. To support these tasks, we construct over 9,000 VQA pairs that systematically evaluate the capabilities of VLMs. Experiments on PlantMicro show that current VLMs struggle with fine-grained recognition and biologically grounded reasoning. For example, GPT-5 achieves 34.93% accuracy on the pathogen classification task, which is only modestly above the random-guessing baseline. The results highlight a significant gap in current VLMs' ability to comprehend plant microscopic images. PlantMicro provides a standardized foundation for advancing VLMs toward reliable and comprehensive microscopy-level plant understanding.

Problem

• vision-language models
• microscopic plant image
• benchmark
• image understanding
• plant pathology

Innovation

• vision-language models
• microscopic plant imaging
• benchmark
• visual question answering
• fine-grained recognition