Cascaded Regression Tracking: Towards Online Hard Distractor Discrimination

TL;DR

This paper introduces a cascaded regression tracking method that effectively discriminates hard distractors in online visual tracking, improving robustness and achieving state-of-the-art results in real-time.

Contribution

The paper proposes a novel two-stage cascaded regression framework that enhances online tracking robustness by effectively filtering out hard distractors using efficient regression techniques.

Findings

Achieves state-of-the-art performance on 11 challenging benchmarks.

Runs in real-time with high accuracy.

Effectively discriminates hard distractors in visual tracking.

Abstract

Visual tracking can be easily disturbed by similar surrounding objects. Such objects as hard distractors, even though being the minority among negative samples, increase the risk of target drift and model corruption, which deserve additional attention in online tracking and model update. To enhance the tracking robustness, in this paper, we propose a cascaded regression tracker with two sequential stages. In the first stage, we filter out abundant easily-identified negative candidates via an efficient convolutional regression. In the second stage, a discrete sampling based ridge regression is designed to double-check the remaining ambiguous hard samples, which serves as an alternative of fully-connected layers and benefits from the closed-form solver for efficient learning. Extensive experiments are conducted on 11 challenging tracking benchmarks including OTB-2013, OTB-2015, VOT2018, VOT2019, UAV123, Temple-Color, NfS, TrackingNet, LaSOT, UAV20L, and OxUvA. The proposed method achieves state-of-the-art performance on prevalent benchmarks, while running in a real-time speed.