Exploring the Underwater World Segmentation without Extra Training

TL;DR

This paper introduces AquaOV255, a large underwater segmentation dataset, and UOVSBench, a benchmark for open-vocabulary underwater segmentation, along with Earth2Ocean, a training-free model that transfers terrestrial vision-language models to underwater scenes.

Contribution

The paper presents the first large-scale underwater segmentation dataset, a comprehensive benchmark, and a novel training-free model for underwater segmentation using terrestrial models.

Findings

Earth2Ocean improves segmentation performance on underwater datasets.

AquaOV255 covers 255 categories with over 20K images.

UOVSBench enables comprehensive evaluation of open-vocabulary underwater segmentation.

Abstract

Accurate segmentation of marine organisms is vital for biodiversity monitoring and ecological assessment, yet existing datasets and models remain largely limited to terrestrial scenes. To bridge this gap, we introduce \textbf{AquaOV255}, the first large-scale and fine-grained underwater segmentation dataset containing 255 categories and over 20K images, covering diverse categories for open-vocabulary (OV) evaluation. Furthermore, we establish the first underwater OV segmentation benchmark, \textbf{UOVSBench}, by integrating AquaOV255 with five additional underwater datasets to enable comprehensive evaluation. Alongside, we present \textbf{Earth2Ocean}, a training-free OV segmentation framework that transfers terrestrial vision--language models (VLMs) to underwater domains without any additional underwater training. Earth2Ocean consists of two core components: a Geometric-guided Visual Mask Generator (\textbf{GMG}) that refines visual features via self-similarity geometric priors for local structure perception, and a Category-visual Semantic Alignment (\textbf{CSA}) module that enhances text embeddings through multimodal large language model reasoning and scene-aware template construction. Extensive experiments on the UOVSBench benchmark demonstrate that Earth2Ocean achieves significant performance improvement on average while maintaining efficient inference.