Context-Aware Streaming Perception in Dynamic Environments

TL;DR

This paper introduces Octopus, a method that dynamically adapts vision system configurations to maximize streaming accuracy in real-time environments, accounting for scene difficulty and obstacle movement.

Contribution

It proposes a novel approach to optimize streaming accuracy for each environment context, improving real-time perception performance in dynamic scenes.

Findings

Improves tracking performance (S-MOTA) by 7.4% over static methods.

Maximizes accuracy in streaming settings by considering scene difficulty and obstacle displacement.

Enhances real-time vision systems without compromising offline accuracy.

Abstract

Efficient vision works maximize accuracy under a latency budget. These works evaluate accuracy offline, one image at a time. However, real-time vision applications like autonomous driving operate in streaming settings, where ground truth changes between inference start and finish. This results in a significant accuracy drop. Therefore, a recent work proposed to maximize accuracy in streaming settings on average. In this paper, we propose to maximize streaming accuracy for every environment context. We posit that scenario difficulty influences the initial (offline) accuracy difference, while obstacle displacement in the scene affects the subsequent accuracy degradation. Our method, Octopus, uses these scenario properties to select configurations that maximize streaming accuracy at test time. Our method improves tracking performance (S-MOTA) by 7.4% over the conventional static approach. Further, performance improvement using our method comes in addition to, and not instead of, advances in offline accuracy.