Publications by authors named "Erandi Monterrubio-Martinez"
Article Synopsis
- Marine oil spills are a global concern that need effective tools for response and recovery, leading to the exploration of Deep Learning models for classification and segmentation using Sentinel-1 SAR imagery.
- Researchers created a new dataset and tested 90 configurations of Convolutional Neural Networks (CNNs) for classification, finding that a model with six layers and 32 filters achieved 99% accuracy.
- For segmentation, the U-Net model demonstrated 99% accuracy and 96% Intersection over Union (IoU) with specific configurations, resulting in a proposed framework achieving 95% overall accuracy and 90% IoU.
Water shortage and contamination is a problem worldwide, impacting the human health. This research provides a comprehensive assessment of water quality and its possible impact on public health in San Luis Potosi, a region in Mexico facing critical water challenges. Throughout the study of various pollutant sources, the contaminants were identified and analyzed.
View Article and Find Full Text PDF