BenthiCat: An opti-acoustic dataset for advancing benthic classification and habitat mapping

TL;DR

BenthiCat tackles the scarcity of large, annotated benthic datasets by presenting a comprehensive opti-acoustic dataset collected along the Catalonia coast. It fuses about one million SSS tiles with ~36k pixel-wise annotations, bathymetric DEMs, and ~178k optical images, and introduces a cross-modal association strategy to enable self-supervised learning across modalities. The authors provide raw data, processed mosaics, and open-source tools, establishing train/test splits and baseline benchmarks to standardize evaluation for seafloor classification and multi-sensor fusion. This resource is poised to advance real-time, autonomous seafloor mapping and reproducible cross-modal research in underwater environments.

Abstract

Benthic habitat mapping is fundamental for understanding marine ecosystems, guiding conservation efforts, and supporting sustainable resource management. Yet, the scarcity of large, annotated datasets limits the development and benchmarking of machine learning models in this domain. This paper introduces a thorough multi-modal dataset, comprising about a million side-scan sonar (SSS) tiles collected along the coast of Catalonia (Spain), complemented by bathymetric maps and a set of co-registered optical images from targeted surveys using an autonomous underwater vehicle (AUV). Approximately \num{36000} of the SSS tiles have been manually annotated with segmentation masks to enable supervised fine-tuning of classification models. All the raw sensor data, together with mosaics, are also released to support further exploration and algorithm development. To address challenges in multi-sensor data fusion for AUVs, we spatially associate optical images with corresponding SSS tiles, facilitating self-supervised, cross-modal representation learning. Accompanying open-source preprocessing and annotation tools are provided to enhance accessibility and encourage research. This resource aims to establish a standardized benchmark for underwater habitat mapping, promoting advancements in autonomous seafloor classification and multi-sensor integration.

Figures (16)

• Figure 1: Overview of the surveyed area, divided into 22 sectors, illustrating a portion of the mosaiced SSS transects from sectors N08 and the corresponding DEM.
• Figure 2: Schema of the equipment on board the survey vessel.
• Figure 3: Illustrative example of a portion of an SSS mosaic from sector N08 overlaid with the corresponding benthic habitat interpretation.
• Figure 4: Overview of the optical transects, depicted as white lines, across the surveyed areas in sectors N07 and N08.
• Figure 5: Image of the Girona1000 AUV at port before the start of the optical surveys.