®®®® SIIA Público

Título del libro: International Conference Image And Vision Computing New Zealand
Título del capítulo: Tool-Assisted Annotation of Seafloor Sediment-linked Features Using Weakly Supervised Semantic Segmentation

Autores UNAM:
Autores externos:
Idioma:
Año de publicación:
2025
Palabras clave:

Computer vision; Ecosystems; Iterative methods; Self-supervised learning; Semantic Segmentation; Semantics; Supervised learning; Benthic habitat monitoring; Feature enhancement; Few-shot learning; Pixel-wise labeling; Seafloor imagery; Seafloor sediments; Semantic segmentation; Underwater computer vision; Vision research; Weakly supervised learning; Sediments

Resumen:

Pixel-wise labeling of seafloor imagery is highly time-consuming, limiting the scalability of benthic habitat monitoring. While existing underwater computer vision research has largely focused on visually prominent habitats, sediment-linked benthic features remain underexplored and lack annotated datasets. To address this gap, we propose a tool-assisted annotation framework based on weakly supervised semantic segmentation. The framework follows a three-phase pipeline: feature-based pseudo-mask generation, binary class-specific segmentation, and iterative multiclass segmentation with pseudo-mask expansion and affinity-field regularization. Applied to six ecologically important sediment features, the approach progressively improves pseudo-label quality while reducing reliance on dense expert annotations, providing a scalable solution that can accelerate the annotation of seafloor sediment features. Code is available at: https://github.com/shahrokh1106/seafloor-sediment-segmentation ©2025 IEEE.

Entidades citadas de la UNAM:
Fuente:


ISBN:
9798331586546
Editorial:
IEEE Computer Society

Tipo de producto:
Conference Paper



DOI:
10.1109/IVCNZ67716.2025.11281825