🌊 How accurate can we get in classifying marine debris?
With DenseNet + NN, our models achieve 84% test accuracy, leading the way in automated underwater monitoring.
With DenseNet + NN, our models achieve 84% test accuracy, leading the way in automated underwater monitoring.
🌍 CLEVER’s research on green 6G AI ensures that as networks scale, they remain energy-efficient and secure. Learn how photonics + AI are converging for a sustainable digital future.
📡 Photonic-Aware Neural Networks (PANNs) make real-time cyber defense possible directly at 6G base stations. This means faster mitigation ⚡ and autonomous threat handling.
🧠 CNN architectures like ResNet, VGG19, and DenseNet power our AI-driven monitoring of deep-sea ecosystems. These tools help us protect oceans through automated, precise debris detection.
The future of 6G networks demands intelligent and sustainable security solutions. Our latest research, “Photonic-accelerated AI for cybersecurity in sustainable 6G networks,” introduces a novel approach using photonic-based Convolutional Neural Networks (CNNs) to detect Denial of Service (DoS) Hulk attacks…
🌊 Tackling marine debris with AI! Our research compares machine learning models (NN, SVM, XGB, LR, RF) for underwater debris classification. Neural Networks and SVM lead with up to 84% accuracy.
⚡ Photonic Convolutional Neural Networks (CNNs) outperform electronics in speed, density, and energy use — making them essential for green 6G security.
🌐 Future networks demand security and efficiency. Our work on Photonic-accelerated AI shows how next-gen photonic hardware can detect cyberattacks in real-time, with 99.7% accuracy.
💡 Prototype in action: CATER seamlessly integrates with Apache Ozone, using modular APIs for intelligent node selection. This ensures flexibility across diverse edge environments like smart factories and hospitals.
🔍 Optimizing edge storage isn’t just about speed — it’s about sustainability. Our CATER framework strikes the balance between optimal performance and real-world deployment, reducing costly data movement while improving efficiency.