We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Marine Debris Detection in Satellite Surveillance Using Attention Mechanisms
Summary
Researchers developed an approach combining satellite imagery with attention-based deep learning models to detect marine debris from space. The study found that a model using combined spatial and channel attention (CBAM) achieved the best performance, with a 77% detection score, outperforming other approaches tested. These findings suggest that AI-enhanced satellite surveillance could become a practical tool for monitoring ocean plastic pollution at scale.
Marine debris poses a critical threat to environmental ecosystems, necessitating effective methods for its detection and localization. This study addresses the existing limitations in the literature by proposing an innovative approach that combines the instance segmentation capabilities of YOLOv7 with various attention mechanisms to enhance efficiency and broaden applicability. The primary contribution lies in the exploration and comparison of three attentional models: lightweight coordinate attention, CBAM (combining spatial and channel focus), and bottleneck transformer based on self-attention. Leveraging a meticulously labeled dataset of satellite images containing ocean debris, the study conducts a comprehensive assessment of box detection and mask evaluation. The results demonstrate that CBAM emerges as the standout performer, achieving the highest F1 score (77%) in box detection, surpassing coordinate attention (71%) and YOLOv7/bottleneck transformer (both around 66%). In mask evaluation, CBAM continues to lead with an F1 score of 73%, while coordinate attention and YOLOv7 exhibit comparable performances (around F1 scores of 68% and 69%), and bottleneck transformer lags behind at an F1 score of 56%. This compelling evidence underscores CBAM's superior suitability for detecting marine debris compared to existing methods. Notably, the study reveals an intriguing aspect of the bottleneck transformer, which, despite lower overall performance, successfully detected areas overlooked by manual annotation. Moreover, it demonstrated enhanced mask precision for larger debris pieces, hinting at potentially superior practical performance in certain scenarios. This nuanced finding underscores the importance of considering specific application requirements when selecting a detection model, as the bottleneck transformer may offer unique advantages in certain contexts.
Sign in to start a discussion.
More Papers Like This
Large-scale detection of marine debris in coastal areas with Sentinel-2
Researchers built a deep learning model to detect floating marine debris in coastal areas using satellite imagery from the Sentinel-2 program. The system achieved strong detection accuracy across multiple test sites and can monitor large stretches of coastline regularly. The tool could help environmental agencies track and respond to marine plastic pollution at a scale that manual surveys cannot match.
A Combination of Machine Learning Algorithms for Marine Plastic Litter Detection Exploiting Hyperspectral PRISMA Data
Researchers applied a combination of machine learning algorithms to hyperspectral satellite imagery from the PRISMA satellite to detect marine plastic litter along coastlines and ocean surfaces. The multi-algorithm approach improved detection accuracy over single-model methods and demonstrated the potential for satellite-based monitoring of ocean plastic pollution at scale.
Deep-Feature-Based Approach to Marine Debris Classification
This study applied deep learning to classify marine debris from images, demonstrating that feature-based neural network approaches can effectively distinguish plastic types and other debris categories to support automated ocean monitoring.
Towards Detecting Floating Objects on a Global Scale with Learned Spatial Features Using Sentinel 2
Researchers developed a machine learning approach using Sentinel-2 satellite imagery to detect floating plastic debris and marine litter on a global scale, demonstrating that learned spatial features can improve detection of large aggregations of floating objects on water surfaces.
AI for Monitoring Ocean Plastic Pollution
This review assessed how artificial intelligence technologies—including satellite image analysis, computer vision, and machine learning—are being applied to monitor ocean plastic pollution. The authors found AI can dramatically expand spatial coverage and detection speed compared to traditional ship-based surveys, though ground-truth validation and data standardization remain challenges.