We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
MSD-AttSegNet: Multiscale Dilated Depthwise Convolution and Channel–Spatial Attention integrated UNet for Microplastic Segmentation in SEM Images
AI summary Read the abstract
Scientists have developed an AI tool that can automatically spot and outline tiny plastic particles in microscope images with over 95% accuracy, outperforming previous detection methods. This matters because microplastics are showing up everywhere, in our water, food, and even our bodies, and faster, more reliable detection tools could help researchers better track contamination levels and understand potential health risks sooner.
Microplastic contamination has become a serious global environmental issue, with synthetic polymer fragments smaller than 5 mm detected across marine, freshwater, terrestrial, and atmospheric ecosystems. Monitoring and identifying macroplastic waste becomes more difficult as it breaks down into microplastics and even nanoplastics due to mechanical, chemical, and biological deterioration. This study proposes a deep learning architecture, MSD-AttSegNet, a hybrid framework that combines multiscale dilated depthwise convolutions and channel–spatial attention with an enhanced UNet. The pipeline integrates a standardized pre-processing strategy, pixel-wise loss functions, and mathematically defined evaluation metrics. A well-known, publicly available Mendeley dataset consisting of 237 micrographs of microplastic particles (fragments or beads) in the range of 50 $$\mu $$ m – 1 mm and fibres with diameters around 10 $$\mu $$ m is used to train the proposed MSD-AttSegNet using k-fold cross-validation. The performance of the model is evaluated by calculating Accuracy (95.51%), Loss (0.57224), Sensitivity (95.56%), Specificity (93.83%), Dice (0.94714), and Jaccard coefficient (0.90). The experiment’s outcome is compared with those of UNet++, TransUNet, LiDSCUNet++, and LiTransUNet. The results indicate improved average performance of the proposed method compared with the evaluated baseline models.
More Papers Like This
Microplastic Contamination Detection in Water using Deep Learning Models
AI summary Read the abstract
Scientists trained a computer program (using AI image recognition) to automatically spot tiny plastic particles in water samples, and it correctly identified microplastics 99.5% of the time—faster and easier than the manual microscope checks experts currently rely on. This matters because microplastics in our water supply are linked to potential health risks, so having a quick, accurate way to detect them could help water treatment facilities catch contamination sooner and keep drinking water safer.
Automated Quantification of Fibrous Microplastics Using Attention Meta U-Net with Advanced Image Processing
AI summary Read the abstract
Scientists have developed an AI-powered tool that automatically counts and measures tiny plastic fibers (from things like synthetic clothing) shed into water, using smart image-scanning technology instead of slow, error-prone manual counting. This matters because these fibers—too small to see easily—can end up in our water and food supply, and having a faster, more accurate way to track them helps researchers understand and eventually reduce our exposure to this widespread pollution.
MicroNet: Automated Classification of Holographic Microplastics and Environmental Impact Analysis using Convolutional Neural Networks and Large Language Models
AI summary Read the abstract
Scientists built an AI tool that can quickly identify different types of microplastic pollution in water samples using special camera images, achieving over 93% accuracy—much faster than traditional lab methods that require expert analysis. This matters because microplastics are increasingly found in our oceans, food, and even our bodies, and faster detection tools could help researchers and environmental agencies track pollution sources and respond more quickly to protect ecosystems (and ultimately, the food chain we depend on).
Automated Quantification of Fibrous Microplastics Using Attention Meta U-Net with Advanced Image Processing
AI summary Read the abstract
Scientists developed a new AI-powered tool that automatically detects and measures tiny plastic fibers—shed from clothes and textiles—in water samples with about 98% accuracy, replacing slow, error-prone manual counting methods. This matters because these microplastic fibers end up in our water, food, and environment, and better detection tools help researchers track pollution levels more reliably, which is a key step toward understanding and eventually reducing our exposure to these particles.
Microplastic Identification via Raman Spectroscopy: A Three‐Channel Feature Fusion Convolutional Neural Network
AI summary Read the abstract
Scientists have developed a smarter AI tool that can identify tiny plastic particles with 99% accuracy, even in messy real-world samples like water or soil where other detection methods often fail. This matters because microplastics are increasingly found in our food, water, and even our bodies, and better detection tools help researchers track pollution levels and understand our actual exposure—an important step toward figuring out what these particles might mean for our health.
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.