Model Decay in Long-Term Tracking

TL;DR

This paper investigates the problem of model decay in long-term visual tracking, analyzing its causes and proposing simple methods to mitigate bias accumulation, leading to improved tracking accuracy and robustness over extended durations.

Contribution

It provides a mathematical analysis of model decay, examines its sources, and introduces simple techniques to reduce bias in long-term tracking.

Findings

Proposed methods improve long-term tracking accuracy.

Enhanced robustness in videos up to 30 minutes long.

Mathematical insights into bias accumulation in trackers.

Abstract

Updating the tracker model with adverse bounding box predictions adds an unavoidable bias term to the learning. This bias term, which we refer to as model decay, offsets the learning and causes tracking drift. While its adverse affect might not be visible in short-term tracking, accumulation of this bias over a long-term can eventually lead to a permanent loss of the target. In this paper, we look at the problem of model bias from a mathematical perspective. Further, we briefly examine the effect of various sources of tracking error on model decay, using a correlation filter (ECO) and a Siamese (SINT) tracker. Based on observations and insights, we propose simple additions that help to reduce model decay in long-term tracking. The proposed tracker is evaluated on four long-term and one short term tracking benchmarks, demonstrating superior accuracy and robustness, even in 30 minute long videos.