EdgeMA: Model Adaptation System for Real-Time Video Analytics on Edge Devices
This addresses the challenge of maintaining accuracy for video analytics on resource-constrained edge devices when faced with domain shifts, representing an incremental improvement in model adaptation techniques.
The paper tackles the problem of data drift in real-time video analytics on edge devices by introducing EdgeMA, a system that adapts models to changing scenes, resulting in significantly improved inference accuracy as demonstrated on a real-world dataset.
Real-time video analytics on edge devices for changing scenes remains a difficult task. As edge devices are usually resource-constrained, edge deep neural networks (DNNs) have fewer weights and shallower architectures than general DNNs. As a result, they only perform well in limited scenarios and are sensitive to data drift. In this paper, we introduce EdgeMA, a practical and efficient video analytics system designed to adapt models to shifts in real-world video streams over time, addressing the data drift problem. EdgeMA extracts the gray level co-occurrence matrix based statistical texture feature and uses the Random Forest classifier to detect the domain shift. Moreover, we have incorporated a method of model adaptation based on importance weighting, specifically designed to update models to cope with the label distribution shift. Through rigorous evaluation of EdgeMA on a real-world dataset, our results illustrate that EdgeMA significantly improves inference accuracy.