ROMA: Run-Time Object Detection To Maximize Real-Time Accuracy

TL;DR

This paper introduces ROMA, a run-time model that dynamically selects the best detector among multiple YOLOv4 models to maximize real-time object detection accuracy under varying video content and latency conditions.

Contribution

ROMA is a novel run-time accuracy variation model that optimally selects detectors in real time without label information, improving accuracy in dynamic scenarios.

Findings

ROMA improves real-time detection accuracy by 4-37%.

ROMA outperforms individual detectors and existing runtime techniques.

The model is validated on NVIDIA Jetson Nano with MOT datasets.

Abstract

This paper analyzes the effects of dynamically varying video contents and detection latency on the real-time detection accuracy of a detector and proposes a new run-time accuracy variation model, ROMA, based on the findings from the analysis. ROMA is designed to select an optimal detector out of a set of detectors in real time without label information to maximize real-time object detection accuracy. ROMA utilizing four YOLOv4 detectors on an NVIDIA Jetson Nano shows real-time accuracy improvements by 4 to 37% for a scenario of dynamically varying video contents and detection latency consisting of MOT17Det and MOT20Det datasets, compared to individual YOLOv4 detectors and two state-of-the-art runtime techniques.