Energy-Efficient Fast Object Detection on Edge Devices for IoT Systems

TL;DR

This work tackles the challenge of energy-efficient, real-time object detection for IoT edge devices by marrying frame-difference motion detection with a lightweight AI classifier. The approach is implemented and evaluated on three edge platforms (Hailo-8, Jetson Orin Nano, and AMD Alveo U50) using four architectures (MobileNet, Inception-V4, ResNet50, ViT Base) across four object classes, comparing against end-to-end detectors like YOLOX. Results show substantial gains: average accuracy up by 28.314%, efficiency by 3.6x, and latency reduced by 39.305% relative to end-to-end, with MobileNet consistently delivering the best balance of accuracy, speed, and energy use. The findings support deploying frame-difference–based FMOD with lightweight classifiers for fast-moving object detection in IoT scenarios where power and latency are critical, especially for trains and airplanes.

Abstract

This paper presents an Internet of Things (IoT) application that utilizes an AI classifier for fast-object detection using the frame difference method. This method, with its shorter duration, is the most efficient and suitable for fast-object detection in IoT systems, which require energy-efficient applications compared to end-to-end methods. We have implemented this technique on three edge devices: AMD AlveoT M U50, Jetson Orin Nano, and Hailo-8T M AI Accelerator, and four models with artificial neural networks and transformer models. We examined various classes, including birds, cars, trains, and airplanes. Using the frame difference method, the MobileNet model consistently has high accuracy, low latency, and is highly energy-efficient. YOLOX consistently shows the lowest accuracy, lowest latency, and lowest efficiency. The experimental results show that the proposed algorithm has improved the average accuracy gain by 28.314%, the average efficiency gain by 3.6 times, and the average latency reduction by 39.305% compared to the end-to-end method. Of all these classes, the faster objects are trains and airplanes. Experiments show that the accuracy percentage for trains and airplanes is lower than other categories. So, in tasks that require fast detection and accurate results, end-to-end methods can be a disaster because they cannot handle fast object detection. To improve computational efficiency, we designed our proposed method as a lightweight detection algorithm. It is well suited for applications in IoT systems, especially those that require fast-moving object detection and higher accuracy.

Figures (12)

• Figure 1: Flow-Chart of Proposed Algorithm
• Figure 2: Frame Difference Process
• Figure 3: Image Dilation Process
• Figure 4: Image Erosion Process
• Figure 5: Image Blurring Process