Mez: A Messaging System for Latency-Sensitive Multi-Camera Machine Vision at the IoT Edge

TL;DR

Mez is a specialized messaging system designed for latency-sensitive multi-camera machine vision at the IoT Edge, dynamically balancing latency and accuracy through adaptive video quality adjustments.

Contribution

It introduces a novel network latency controller and domain-specific optimizations for low-latency, accuracy-aware messaging in IoT edge machine vision applications.

Findings

Tolerates latency variations up to 10x

Achieves only 4.2% reduction in F1 score under worst-case conditions

Demonstrates effectiveness on an IoT Edge testbed with pedestrian detection

Abstract

Mez is a publish-subscribe messaging system for latency sensitive multi-camera machine vision at the IoT Edge. Unlike existing messaging systems, Mez allows applications to specify latency, and application accuracy bounds. Mez implements a network latency controller that dynamically adjusts the video frame quality to satisfy latency, and application accuracy requirements. Additionally, the design of Mez utilizes application domain specific features to provide low latency operations. Experimental evaluation on an IoT Edge testbed with a pedestrian detection machine vision application indicates that Mez is able to tolerate latency variations of up to 10x with a worst-case reduction of 4.2\% in the application accuracy F1 score metric.