Motion Vector Extrapolation for Video Object Detection

TL;DR

This paper introduces MOVEX, a method combining off-the-shelf object detectors with optical flow to significantly reduce latency in video object detection without losing accuracy, enabling efficient CPU-based processing.

Contribution

MOVEX presents a novel approach that leverages existing detectors and optical flow for low-latency video object detection, surpassing current methods in speed while maintaining accuracy.

Findings

Reduces detection latency up to 25x on MOT20 dataset

Maintains accuracy comparable to baseline detectors

Enables CPU-based real-time video object detection

Abstract

Despite the continued successes of computationally efficient deep neural network architectures for video object detection, performance continually arrives at the great trilemma of speed versus accuracy versus computational resources (pick two). Current attempts to exploit temporal information in video data to overcome this trilemma are bottlenecked by the state-of-the-art in object detection models. We present, a technique which performs video object detection through the use of off-the-shelf object detectors alongside existing optical flow based motion estimation techniques in parallel. Through a set of experiments on the benchmark MOT20 dataset, we demonstrate that our approach significantly reduces the baseline latency of any given object detector without sacrificing any accuracy. Further latency reduction, up to 25x lower than the original latency, can be achieved with minimal accuracy loss. MOVEX enables low latency video object detection on common CPU based systems, thus allowing for high performance video object detection beyond the domain of GPU computing. The code is available at https://github.com/juliantrue/movex.