Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

State of the art applications of deep learning within tracking and detecting marine debris: A survey

Zoe Moorton, Zeyneb Kurt, Wai Lok Woo

TL;DR

This survey addresses the problem of detecting and tracking marine debris with deep learning, focusing on macro-scale debris and remote-sensing as well as in-water imaging approaches. It systematically reviews 28 recent contributions, compares object-detection/classification methods (notably the YOLO family) and datasets (e.g., MARIDA, TrashCan, CleanSea), and highlights the pervasive lack of large, publicly available, diverse debris datasets. The authors synthesize current trends, identify critical data-collection gaps, and propose more than 40 future directions, including data augmentation, synthetic data generation, multi-sensor fusion, and robust evaluation protocols, to accelerate practical debris monitoring and cleanup. Collectively, the work provides a comprehensive pointer for researchers and practitioners aiming to deploy deep learning for ocean-cleanup applications and river-debris interception, with implications for policy, conservation, and automated robotics.

Abstract

Deep learning techniques have been explored within the marine litter problem for approximately 20 years but the majority of the research has developed rapidly in the last five years. We provide an in-depth, up to date, summary and analysis of 28 of the most recent and significant contributions of deep learning in marine debris. From cross referencing the research paper results, the YOLO family significantly outperforms all other methods of object detection but there are many respected contributions to this field that have categorically agreed that a comprehensive database of underwater debris is not currently available for machine learning. Using a small dataset curated and labelled by us, we tested YOLOv5 on a binary classification task and found the accuracy was low and the rate of false positives was high; highlighting the importance of a comprehensive database. We conclude this survey with over 40 future research recommendations and open challenges.

State of the art applications of deep learning within tracking and detecting marine debris: A survey

TL;DR

This survey addresses the problem of detecting and tracking marine debris with deep learning, focusing on macro-scale debris and remote-sensing as well as in-water imaging approaches. It systematically reviews 28 recent contributions, compares object-detection/classification methods (notably the YOLO family) and datasets (e.g., MARIDA, TrashCan, CleanSea), and highlights the pervasive lack of large, publicly available, diverse debris datasets. The authors synthesize current trends, identify critical data-collection gaps, and propose more than 40 future directions, including data augmentation, synthetic data generation, multi-sensor fusion, and robust evaluation protocols, to accelerate practical debris monitoring and cleanup. Collectively, the work provides a comprehensive pointer for researchers and practitioners aiming to deploy deep learning for ocean-cleanup applications and river-debris interception, with implications for policy, conservation, and automated robotics.

Abstract

Deep learning techniques have been explored within the marine litter problem for approximately 20 years but the majority of the research has developed rapidly in the last five years. We provide an in-depth, up to date, summary and analysis of 28 of the most recent and significant contributions of deep learning in marine debris. From cross referencing the research paper results, the YOLO family significantly outperforms all other methods of object detection but there are many respected contributions to this field that have categorically agreed that a comprehensive database of underwater debris is not currently available for machine learning. Using a small dataset curated and labelled by us, we tested YOLOv5 on a binary classification task and found the accuracy was low and the rate of false positives was high; highlighting the importance of a comprehensive database. We conclude this survey with over 40 future research recommendations and open challenges.
Paper Structure (27 sections, 1 figure, 2 tables)

This paper contains 27 sections, 1 figure, 2 tables.

Table of Contents

  1. Introduction
  2. Background
  3. The Rise of Plastic
  4. The Plastic Cycle
  5. A Brief History of AI in Marine Debris
  6. Timeline
  7. Deep Learning for Other Marine Applications
  8. Deep learning techniques in marine debris identification
  9. Remote Sensing
  10. Classification & Object Detection
  11. Deep Learning for River Debris Identification
  12. Alternative applications
  13. Computational Cost and Performance
  14. Data Acquisition
  15. Benchmark Databases
  16. ...and 12 more sections

Figures (1)

  • Figure 1: Outcome of highest accuracy test. Figures A and B (Top Row): Animal. Figures C and D (Bottom Row): Litter.