# Consistency-Driven Dual-Teacher Framework for Semi-Supervised Zooplankton Microscopic Image Segmentation

## TL;DR

This paper introduces a new framework for segmenting zooplankton images using two teacher networks to improve accuracy with limited labeled data.

## Contribution

A dual-teacher framework with dynamic pseudo-label filtering is proposed for zooplankton image segmentation.

## Key findings

- The framework outperforms existing methods with mIoU scores up to 73.92% using only half the labeled data.
- Dynamic fusion of pseudo-labels improves performance across different annotation ratios.
- The method is tested on a newly constructed microscopic zooplankton dataset.

## Abstract

In-depth research on marine biodiversity is essential for understanding and protecting marine ecosystems, where semantic segmentation of marine species plays a crucial role. However, segmenting microscopic zooplankton images remains challenging due to highly variable morphologies, complex boundaries, and the scarcity of high-quality pixel-level annotations that require expert knowledge. Existing semi-supervised methods often rely on single-model perspectives, producing unreliable pseudo-labels and limiting performance in such complex scenarios. To address these challenges, this paper proposes a consistency-driven dual-teacher framework tailored for zooplankton segmentation. Two heterogeneous teacher networks are employed: one captures global morphological features, while the other focuses on local fine-grained details, providing complementary and diverse supervision and alleviating overfitting under limited annotations. In addition, a dynamic fusion-based pseudo-label filtering strategy is introduced to adaptively integrate hard and soft labels by jointly considering prediction consistency and confidence scores, thereby enhancing supervision flexibility. Extensive experiments on the Zooplankton-21 Microscopic Segmentation Dataset (ZMS-21), a self-constructed microscopic zooplankton dataset demonstrate that the proposed method consistently outperforms existing semi-supervised segmentation approaches under various annotation ratios, achieving mIoU scores of 64.80%, 69.58%, 70.32%, and 73.92% with 1/16, 1/8, 1/4, and 1/2 labeled data, respectively.