arXiv:2510.18513v2 Announce Type: replace 
Abstract: The rise of convenience packaging has led to generation of enormous waste, making efficient waste sorting crucial for sustainable waste management. To address this, we developed DWaste, a computer vision-powered platform designed for real-time waste sorting on resource-constrained smartphones and edge devices, including offline functionality. We benchmarked various image classification models (EfficientNetV2S/M, ResNet50/101, MobileNet) and object detection (YOLOv8n, YOLOv11n) including our purposed YOLOv8n-CBAM model using our annotated dataset designed for recycling. We found a clear trade-off between accuracy and resource consumption: the best classifier, EfficientNetV2S, achieved high accuracy(~ 96%) but suffered from high latency (~ 0.22s) and elevated carbon emissions. In contrast, lightweight object detection models delivered strong performance (up to 80% mAP) with ultra-fast inference (~ 0.03s) and significantly smaller model sizes (< 7MB ), making them ideal for real-time, low-power use. Model quantization further maximized efficiency, substantially reducing model size and VRAM usage by up to 75%. Our work demonstrates the successful implementation of "Greener AI" models to support real-time, sustainable waste sorting on edge devices.

DWaste هي منصة مبتكرة تستخدم رؤية الكمبيوتر لتحسين فرز النفايات باستخدام الأجهزة المحمولة وأجهزة الحافة. هذا التطور مهم لأنه يعالج المشكلة المتزايدة للنفايات الناتجة عن التعبئة المريحة، مما يعزز ممارسات إدارة النفايات المستدامة. من خلال تمكين الفرز في الوقت الحقيقي على الأجهزة ذات الموارد المحدودة، لا تعمل DWaste على تحسين الكفاءة فحسب، بل تجعل أيضًا إدارة النفايات متاحة لجمهور أوسع، مما قد يؤدي إلى بيئة أنظف.

DWaste es una plataforma innovadora que utiliza visión por computadora para mejorar la clasificación de residuos mediante dispositivos móviles y edge. Este desarrollo es significativo ya que aborda el creciente problema de los residuos generados por el embalaje de conveniencia, promoviendo prácticas de gestión de residuos sostenibles. Al permitir la clasificación en tiempo real en dispositivos con recursos limitados, DWaste no solo mejora la eficiencia, sino que también hace que la gestión de residuos sea accesible a un público más amplio, lo que podría llevar a un entorno más limpio.

DWaste est une plateforme innovante qui utilise la vision par ordinateur pour améliorer le tri des déchets à l'aide de dispositifs mobiles et edge. Ce développement est important car il répond au problème croissant des déchets générés par les emballages pratiques, favorisant des pratiques de gestion des déchets durables. En permettant un tri en temps réel sur des appareils à ressources limitées, DWaste améliore non seulement l'efficacité, mais rend également la gestion des déchets accessible à un plus large public, ce qui pourrait conduire à un environnement plus propre.

DWaste is an innovative platform that leverages computer vision to enhance waste sorting using mobile and edge devices. This development is significant as it addresses the growing issue of waste generated by convenience packaging, promoting sustainable waste management practices. By enabling real-time sorting on resource-constrained devices, DWaste not only improves efficiency but also makes waste management accessible to a wider audience, potentially leading to a cleaner environment.

DWaste: Greener AI for Waste Sorting using Mobile and Edge Devices

arXiv:2601.07957v1 Announce Type: new 
Abstract: Maize disease classification plays a vital role in mitigating yield losses and ensuring food security. However, the deployment of traditional disease detection models in resource-constrained environments, such as those using smartphones and drones, faces challenges due to high computational costs. To address these challenges, we propose LWMSCNN-SE, a lightweight convolutional neural network (CNN) that integrates multi-scale feature extraction, depthwise separable convolutions, and squeeze-and-Excitation (SE) attention mechanisms. This novel combination enables the model to achieve 96.63% classification accuracy with only 241,348 parameters and 0.666 GFLOPs, making it suitable for real-time deployment in field applications. Our approach addresses the accuracy--efficiency trade-off by delivering high accuracy while maintaining low computational costs, demonstrating its potential for efficient maize disease diagnosis on edge devices in precision farming systems.

LWMSCNN-SE هو شبكة عصبية تلافيفية خفيفة الوزن مقترحة حديثًا لتصنيف أمراض الذرة بكفاءة، حيث تحقق دقة تصل إلى 96.63% مع متطلبات حسابية منخفضة، مما يجعلها مناسبة للنشر على الأجهزة الطرفية مثل الهواتف الذكية والطائرات بدون طيار.

LWMSCNN-SE es una nueva red neuronal convolucional ligera propuesta para la clasificación eficiente de enfermedades del maíz, logrando una precisión del 96.63 % con requisitos computacionales mínimos, lo que la hace adecuada para su implementación en dispositivos de borde como teléfonos inteligentes y drones.

LWMSCNN-SE est un nouveau réseau de neurones convolutifs léger proposé pour la classification efficace des maladies du maïs, atteignant une précision de 96,63 % avec des exigences computationnelles minimales, ce qui le rend adapté au déploiement sur des appareils en périphérie tels que les smartphones et les drones.

LWMSCNN-SE is a newly proposed lightweight convolutional neural network designed for efficient maize disease classification, achieving 96.63% accuracy with minimal computational requirements, making it suitable for deployment on edge devices like smartphones and drones.

LWMSCNN-SE: A Lightweight Multi-Scale Network for Efficient Maize Disease Classification on Edge Devices

One More Thing in AI – Your Shortcut to AI Mastery

DWaste: Greener AI for Waste Sorting using Mobile and Edge Devices

Was this article worth reading? Share it

One More Thing in AI

Bulk Image Generation AI

Recereum

Solvice

MindStudio

Ronin8 Technologies

Ready to build your own newsroom?