Deep learning-based underwater species recognition: Advances, benchmarks and challenges

Chen Guo

doi:10.1016/j.image.2026.117562

Deep learning-based underwater species recognition: Advances, benchmarks and challenges

2026-04-21

0

PRE

AI

C

Chen Guo

W

Weiling Chen

T

Tiesong Zhao *

DOI：10.1016/j.image.2026.117562

Original

Original request for help

Save

Abstract

En 中文

• A comprehensive review of recent deep learning-based underwater species recognition methods. • A systematic analysis of fine-grained and long-tailed challenges in underwater scenarios. • A unified benchmark analysis with large-scale datasets and comparative evaluation of representative methods. • A discussion of emerging research directions for underwater species recognition.

Keywords:

underwater species recognition

deep learning

fine-grained recognition

long-tailed distribution

benchmark analysis

Deep learning-based underwater species recognition: Advances, benchmarks and challenges

Abstract

Journal

Organization