We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Towards More Efficient EfficientDets and Real-Time Marine Debris Detection
AI summary Read the abstract
Researchers improved the efficiency of a class of AI-based object detection systems called EfficientDets for real-time identification of marine debris underwater. Their optimized models achieved better accuracy while running faster, making them more practical for use on autonomous underwater vehicles. This technology could help enable automated detection and removal of ocean plastic waste, which breaks down into harmful microplastics over time.
Marine debris is a problem both for the health of marine environments and for the human health since tiny pieces of plastic called “microplastics” resulting from the debris decomposition over the time are entering the food chain at any levels. For marine debris detection and removal, autonomous underwater vehicles (AUVs) are a potential solution. In this letter, we focus on the efficiency of AUV vision for real-time marine debris detection. First, we improved the efficiency of a class of state-of-the-art object detectors, namely EfficientDets [1], by 1.5% AP on D0, 2.6% AP on D1, 1.2% AP on D2 and 1.3% AP on D3 without increasing the GPU latency (see Fig. 1 ). Subsequently, we created and made publicly available a dataset for the detection of in-water plastic bags and bottles and trained our improved EfficientDets on this and on two public datasets for marine debris detection. Finally, we began the testing of real-time detection performance on a simulator of marine environments.
More Papers Like This
Towards More Efficient EfficientDets and Low-Light Real-Time Marine Debris Detection
AI summary Read the abstract
Researchers improved the computational efficiency of EfficientDet object detectors and applied them to real-time marine debris detection for autonomous underwater vehicles, achieving AP gains of 1.2-2.6% without increased GPU latency, creating a new in-water plastic bag and bottle dataset, and investigating detection performance degradation under low-light underwater conditions.
Efficient Object Detection of Marine Debris using Pruned YOLO Model
AI summary Read the abstract
Researchers applied a pruned YOLOv4 object detection model to identify marine debris, including microplastic-contaminated items, in underwater images. Channel pruning reduced the model size by 90% while maintaining detection accuracy, making it suitable for deployment on autonomous underwater vehicles. The study demonstrates that efficient machine learning models can support real-time marine debris detection and cleanup efforts.
LE-RTDETR: multi-scale edge enhancement and structural compression for underwater debris detection
AI summary Read the abstract
Researchers built a faster, lighter AI tool that spots trash in murky underwater photos more accurately than previous models, even with blurry visibility and distorted colors. This matters because better detection of ocean debris, including plastic waste that breaks down into microplastics, could help cleanup efforts target pollution that eventually affects marine life and the seafood we eat.
AI – Driven Marine Debris Detection for Ocean Conservation
AI summary Read the abstract
Researchers developed an AI-driven marine debris detection system using the YOLOv8 deep learning model trained to identify plastic waste in challenging underwater conditions including low visibility and complex backgrounds. The system aims to provide scalable, automated monitoring to support ocean conservation and guide debris removal efforts.
Deep Learning-Based Underwater Trash Detection System UsingYOLOv8
AI summary Read the abstract
Researchers trained a YOLOv8 deep learning model on underwater debris imagery to enable real-time detection and classification of submerged plastic trash, demonstrating substantially faster and more scalable performance than traditional manual survey or sonar-based methods.
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.