0
Article Tier 2 Environmental Sources Human Health Effects Marine & Wildlife Sign in to save

SAM-Augmented Blending for Enhanced Microplastic Detection Using YOLO11

2025

AI summary Read the abstract

Researchers developed a synthetic data augmentation method using SAM-generated instance masks combined with the YOLOv11 object detection architecture to improve underwater microplastic detection in images. The approach significantly improved detection performance for small, sparsely labeled microplastic objects where real training data is scarce.

Underwater microplastic pollution poses a significant threat to marine ecosystems and human health, yet accurate detection remains challenging due to the scarcity of annotated data and the small size of target objects. This study presents a synthetic data augmentation approach to enhance underwater microplastic detection using the YOLOv11 object detection architecture. Instance masks are generated with the Segment Anything Model (SAM), enabling the creation of synthetic training data through blending and copy-paste techniques on diverse aquatic backgrounds. We train and evaluate five YOLOv11 models on a consistent real-world test set, demonstrating that all synthetic variants outperform the baseline model, with the SAM-blended (1BG) model achieving the highest mAP@0.5:0.95. These results highlight the effectiveness of SAM-guided synthetic augmentation in addressing data limitations and improving performance for underwater object detection tasks.

More Papers Like This

Article Tier 2

A Fully Automated YOLO-SAM Framework for High-Precision Microplastic Analysis in Environmental Sample

AI summary Read the abstract

Researchers built an automated computer program that combines two AI tools to detect and outline microplastic particles in microscope images. Tested on 5,653 images, the fully automatic system matched particle shapes accurately, and performed even better when given some manual guidance, outperforming an existing method by more than 22%.

Article

Uso de YOLOv8 y aumento de datos para la detección de microplásticos en agua

AI summary Read the abstract

Researchers trained a YOLOv8 computer vision model on 500 stereoscopic microscope images of microplastics collected from two Mexican water bodies, finding that applying data augmentation techniques improved detection precision to 0.825 and mAP@50 to 0.73, outperforming training without augmentation.

Article Tier 2

GANsemble for Small and Imbalanced Data Sets: A Baseline for Synthetic Microplastics Data

AI summary Read the abstract

Researchers developed a generative adversarial network (GAN) ensemble approach to generate synthetic training data for microplastic image classification, addressing the challenge that real microplastic image datasets are small and imbalanced by polymer type. The synthetic augmentation improved classifier accuracy and recall, particularly for underrepresented plastic categories.

Article Tier 2

Application of the YOLOv5 Model for the Detection of Microobjects in the Marine Environment

AI summary Read the abstract

Researchers trained the YOLOv5 machine learning model to automatically detect and classify microplastics and microplankton in underwater photo and video images in real time. The system achieved accuracy comparable to manual counting, offering a faster, less labor-intensive method for monitoring microplastic contamination in marine environments.

Article Tier 2

Digital holography with deep learning and generative adversarial networks for automatic microplastics classification

AI summary Read the abstract

Researchers combined digital holography with deep learning and generative adversarial networks to automatically classify microplastic particles in water. The system achieves high accuracy even with limited training images, making it a practical tool for automated marine microplastic monitoring.

Research digests by email

When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.

Email me about

Share this paper