Online Domain Adaptation for Multi-Object Tracking

TL;DR

This paper introduces an online domain adaptation method for multi-object tracking that improves detector performance by adapting from generic datasets to specific video domains, enhancing tracking accuracy.

Contribution

It is the first to propose online domain adaptation for object detectors in causal multi-object tracking, using a multi-task learning approach to reduce dataset bias and improve recall.

Findings

Improved tracking performance on KITTI benchmark.

Effective adaptation with both simple and complex features.

Reduced dataset bias through online multi-task learning.

Abstract

Automatically detecting, labeling, and tracking objects in videos depends first and foremost on accurate category-level object detectors. These might, however, not always be available in practice, as acquiring high-quality large scale labeled training datasets is either too costly or impractical for all possible real-world application scenarios. A scalable solution consists in re-using object detectors pre-trained on generic datasets. This work is the first to investigate the problem of on-line domain adaptation of object detectors for causal multi-object tracking (MOT). We propose to alleviate the dataset bias by adapting detectors from category to instances, and back: (i) we jointly learn all target models by adapting them from the pre-trained one, and (ii) we also adapt the pre-trained model on-line. We introduce an on-line multi-task learning algorithm to efficiently share parameters and reduce drift, while gradually improving recall. Our approach is applicable to any linear object detector, and we evaluate both cheap "mini-Fisher Vectors" and expensive "off-the-shelf" ConvNet features. We quantitatively measure the benefit of our domain adaptation strategy on the KITTI tracking benchmark and on a new dataset (PASCAL-to-KITTI) we introduce to study the domain mismatch problem in MOT.