Simple Unsupervised Multi-Object Tracking

Karthik, S.; Prabhu, Ameya; Gandhi, Vineet

doi:10.48550/arxiv.2006.02609

Cited by 21 publications

(36 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“….597 (42) 1647 (16) .608 (38) .634 (32) .717 (39) .586 (28) .440 (25) .360 (27) .502 (15) * CSTrack [38] .749 ( 6) 3567 (90) .721 (6) .726 (6) .830 (6) .627 (13) .450 (20) .356 (28) .429 (56) * CLTSMOT…”

Section: A Extended Results: Mot17mentioning

confidence: 99%

Local Metrics for Multi-Object Tracking

Valmadre,

Bewley,

Huang

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper introduces temporally local metrics for Multi-Object Tracking. These metrics are obtained by restricting existing metrics based on track matching to a finite temporal horizon, and provide new insight into the ability of trackers to maintain identity over time. Moreover, the horizon parameter offers a novel, meaningful mechanism by which to define the relative importance of detection and association, a common dilemma in applications where imperfect association is tolerable. It is shown that the historical Average Tracking Accuracy (ATA) metric exhibits superior sensitivity to association, enabling its proposed local variant, ALTA, to capture a wide range of characteristics. In particular, ALTA is better equipped to identify advances in association independent of detection. The paper further presents an error decomposition for ATA that reveals the impact of four distinct error types and is equally applicable to ALTA. The diagnostic capabilities of ALTA are demonstrated on the MOT 2017 and Waymo Open Dataset benchmarks.

show abstract

Section: A Extended Results: Mot17mentioning

confidence: 99%

Local Metrics for Multi-Object Tracking

Valmadre,

Bewley,

Huang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…To reduce the computation, some works attempt to share the Re-ID feature computation with the backbone in anchor-based detector [30,47,44,34] or point-based detector [55] by introducing an extra Re-ID branch that is parallel to detection branch. The common practice in MOT to train the Re-ID module is to classify each identity into one class [60,22,55,47]. There are two fundamental weaknesses of such methods: 1) the Re-ID module is less scalable especially when the amount of identities is huge, because the classifier takes up a lot of memory.…”

Section: Re-identification For Associationmentioning

confidence: 99%

“…Despite the advance in supervised Re-ID learning [18,27,41,28], some works for unsupervised Re-ID learning have been proposed [17,46,22,55,49,15]. These works can be divided into two categories: pseudo identity based [22,55,49,15] and identity free methods [46]. The proposed method is also an identity free method.…”

Section: Re-identification For Associationmentioning

confidence: 99%

“…The proposed method is also an identity free method. For the former category, pseudo identities can be obtained by clustering [49,15] or tracking [22]. However, the errors may accumulate and it is challenging to estimate the number of pseudo identities while clustering, and a trajectory of an object breaks into several short trajectories easily while tracking.…”

Section: Re-identification For Associationmentioning

confidence: 99%

See 1 more Smart Citation

Online Multi-Object Tracking with Unsupervised Re-Identification Learning and Occlusion Estimation

Liu¹,

Chen²,

Chu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Occlusion between different objects is a typical challenge in Multi-Object Tracking (MOT), which often leads to inferior tracking results due to the missing detected objects. The common practice in multi-object tracking is re-identifying the missed objects after their reappearance. Though tracking performance can be boosted by the re-identification, the annotation of identity is required to train the model. In addition, such practice of re-identification still can not track those highly occluded objects when they are missed by the detector. In this paper, we focus on online multi-object tracking and design two novel modules, the unsupervised reidentification learning module and the occlusion estimation module, to handle these problems. Specifically, the proposed unsupervised re-identification learning module does not require any (pseudo) identity information nor suffer from the scalability issue. The proposed occlusion estimation module tries to predict the locations where occlusions happen, which are used to estimate the positions of missed objects by the detector. Our study shows that, when applied to stateof-the-art MOT methods, the proposed unsupervised re-identification learning is comparable to supervised re-identification learning, and the tracking performance is further improved by the proposed occlusion estimation module.

show abstract

“…Neural network based trackers are shown to benefit from directly differentiable tracking-specific loss functions in [52]. Weakly supervised methods such as SimpleReID [26] have also been successful for tracking tasks.…”

Section: Introductionmentioning

confidence: 99%

Localization-Based Tracking

Gloudemans¹,

Work²

2021

Preprint

View full text Add to dashboard Cite

End-to-end production of object tracklets from high resolution video in real-time and with high accuracy remains a challenging problem due to the cost of object detection on each frame. In this work we present Localization-based Tracking (LBT), an extension to any tracker that follows the tracking by detection or joint detection and tracking paradigms. Localization-based Tracking focuses only on regions likely to contain objects to boost detection speed and avoid matching errors. We evaluate LBT as an extension to two example trackers (KIOU and SORT) on the UA-DETRAC and MOT20 datasets. LBT-extended trackers outperform all other reported algorithms in terms of PR-MOTA, PR-MOTP, and mostly tracked objects on the UA-DETRAC benchmark, establishing a new state-of-the art. Relative to tracking by detection with KIOU, LBT-extended KIOU achieves a 25% higher frame-rate and is 1.1% more accurate in terms of PR-MOTA on the UA-DETRAC dataset. LBT-extended SORT achieves a 62% speedup and a 3.2% increase in PR-MOTA on the UA-DETRAC dataset. On MOT20, LBT-extended KIOU has a 50% higher frame-rate than tracking by detection and is 0.4% more accurate in terms of MOTA. As of submission time, our LBT-extended KIOU tracker places 10th overall on the MOT20 benchmark.

show abstract

Simple Unsupervised Multi-Object Tracking

Cited by 21 publications

References 56 publications

Local Metrics for Multi-Object Tracking

Local Metrics for Multi-Object Tracking

Online Multi-Object Tracking with Unsupervised Re-Identification Learning and Occlusion Estimation

Localization-Based Tracking

Contact Info

Product

Resources

About