This research uses edge computing to overcome real-time processing issues in Internet of Things (IoT) picture identification using Convolutional Neural Networks (CNNs). The main goals are to study lightweight CNN architectures, model compression, and edge computing's impact on latency and bandwidth. The article reviews CNN literature and its use in resource-constrained IoT situations using secondary data. Significant results show that lightweight models like MobileNet and EfficientNet are essential for efficient picture identification without losing accuracy, while edge computing decreases latency and improves real-time decision-making. Hardware accelerators help install complicated CNN models on IoT devices. Limitations include edge infrastructure reliance and data privacy issues. The report stresses the need for solid data protection regulations and energy-efficient hardware research. Policymakers should establish IoT data protection standards and accessible edge infrastructure to ensure fair deployment of CNN-based image recognition systems across varied geographies. This study helps explain CNN-IoT synergy, laying the groundwork for real-time image processing advances.
You may also start an advanced similarity search for this article.