A Real-Time Floating Algae Segmentation Using Distillation Network

Jingjing Xu,Lei Wang

DOI: https://doi.org/10.3390/jmse12060852

IF: 2.744

2024-05-22

Journal of Marine Science and Engineering

Abstract:The segmentation of floating algae is a hot topic in the field of marine environmental research. Given the vastness of coastal areas and complex environments, algae detection models must have both higher performance and lower deployment costs. However, relying solely on a single Convolutional Neural Network (CNN) or transformer structure fails to achieve this objective. In this paper, a novel real-time floating algae segmentation method using a distillation network (ADNet) is proposed, based on the RGB images. ADNet can effectively transfer the performance of the transformer-based teacher network to the CNN-based student model while preserving its lightweight design. Faced with complex marine environments, we introduce a novel Channel Purification Module (CPM) to simultaneously strengthen algae features and purify interference responses. Importantly, the CPM achieves this operation without increasing any learnable parameters. Moreover, considering the huge scale differences among algae targets in surveillance RGB images, we propose a lightweight multi-scale feature fusion network (L-MsFFN) to improve the student's modeling ability across various scales. Additionally, to mitigate interference from low-level noises on higher-level semantics, a novel position purification module (PPM) is proposed. The PPM can achieve more accurate weight attention calculation between different pyramid levels, thereby enhancing the effectiveness of fusion. Compared to CNNs and transformers, our ADNet strikes an optimal balance between performance and speed. Extensive experimental results demonstrate that our ADNet achieves higher application performance in the field of floating algae monitoring tasks.

engineering, ocean,oceanography, marine

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve

This paper primarily aims to address the real-time segmentation of floating algae (such as Ulva and Sargassum) in marine environments. Specifically, the paper proposes a new method based on a distillation network—ADNet, with the following objectives:

1. **Balancing High Accuracy and Low Deployment Cost**:
- Existing floating algae detection models are either highly accurate but expensive or cost-effective but lack precision. ADNet attempts to balance high performance and low cost by combining the advantages of Transformer and CNN.

2. **Robustness in Complex Environments**:
- The marine environment is complex and variable, and traditional methods are easily disrupted. ADNet introduces a Channel Purification Module (CPM) to enhance algae features and purify interference responses without adding extra parameters.

3. **Multi-Scale Feature Fusion Capability**:
- Floating algae exhibit significant scale differences in monitoring images, requiring the network to have strong multi-scale feature fusion capabilities. To this end, the paper proposes a Lightweight Multi-Scale Feature Fusion Network (L-MsFFN) to improve modeling capabilities at different scales.

4. **Improving Semantic Consistency**:
- In a multi-layer pyramid structure, noise in lower-level features can interfere with higher-level semantics. Therefore, the paper introduces a Position Purification Module (PPM) to achieve more accurate weight attention calculations between different pyramid levels, thereby enhancing fusion effects.

Through these innovations, ADNet can efficiently and accurately perform floating algae segmentation tasks in complex marine environments.