FinSight-Net:A Physics-Aware Decoupled Network with Frequency-Domain Compensation for Underwater Fish Detection in Smart Aquaculture

2026-02-22

AI Summary

This work addresses the challenges of underwater fish detection, where wavelength-selective absorption and turbidity-induced scattering lead to reduced contrast, structural blurring, and backscattering noise. To tackle these issues, the authors propose a physics-aware, efficient detection framework featuring a multi-scale decoupled dual-stream processing (MS-DDSP) bottleneck that explicitly models and compensates for frequency-domain information degradation. Additionally, an efficient path-aggregation feature pyramid network (EPA-FPN) is introduced, leveraging long-range skip connections and pruning redundant fusion pathways to effectively restore high-frequency spatial details. Evaluated on the UW-BlurredFish dataset, the model achieves a 92.8% mAP—4.8% higher than YOLOv11s—while reducing parameter count by 29.0%, demonstrating a favorable balance between accuracy and model compactness for real-time fish detection in complex aquaculture environments.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Hardware-aware ML
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Underwater fish detection (UFD) is a core capability for smart aquaculture and marine ecological monitoring. While recent detectors improve accuracy by stacking feature extractors or introducing heavy attention modules, they often incur substantial computational overhead and, more importantly, neglect the physics that fundamentally limits UFD: wavelength-dependent absorption and turbidity-induced scattering significantly degrade contrast, blur fine structures, and introduce backscattering noise, leading to unreliable localization and recognition. To address these challenges, we propose FinSight-Net, an efficient and physics-aware detection framework tailored for complex aquaculture environments. FinSight-Net introduces a Multi-Scale Decoupled Dual-Stream Processing (MS-DDSP) bottleneck that explicitly targets frequency-specific information loss via heterogeneous convolutional branches, suppressing backscattering artifacts while compensating distorted biological cues through scale-aware and channel-weighted pathways. We further design an Efficient Path Aggregation FPN (EPA-FPN) as a detail-filling mechanism: it restores high-frequency spatial information typically attenuated in deep layers by establishing long-range skip connections and pruning redundant fusion routes, enabling robust detection of non-rigid fish targets under severe blur and turbidity. Extensive experiments on DeepFish, AquaFishSet, and our challenging UW-BlurredFish benchmark demonstrate that FinSight-Net achieves state-of-the-art performance. In particular, on UW-BlurredFish, FinSight-Net reaches 92.8% mAP, outperforming YOLOv11s by 4.8% while reducing parameters by 29.0%, providing a strong and lightweight solution for real-time automated monitoring in smart aquaculture.

Problem

• underwater fish detection
• wavelength-dependent absorption
• turbidity-induced scattering
• backscattering noise
• smart aquaculture

Innovation

• physics-aware
• frequency-domain compensation
• decoupled network
• underwater fish detection
• multi-scale processing