117 Hits in 5.2 sec

HOTA: A Higher Order Metric for Evaluating Multi-Object Tracking [article]

Jonathon Luiten, Aljosa Osep, Patrick Dendorfer, Philip Torr, Andreas Geiger, Laura Leal-Taixe, Bastian Leibe
2020 arXiv   pre-print
To address this, we present a novel MOT evaluation metric, HOTA (Higher Order Tracking Accuracy), which explicitly balances the effect of performing accurate detection, association and localization into  ...  Multi-Object Tracking (MOT) has been notoriously difficult to evaluate. Previous metrics overemphasize the importance of either detection or association.  ...  Conclusion In this paper, we introduce HOTA (Higher Order Tracking Accuracy), a novel metric for evaluating multi-object tracking.  ... 
arXiv:2009.07736v2 fatcat:sxoithevmvgy7mke5nk2v6bvoy

HOTA: A Higher Order Metric for Evaluating Multi-object Tracking

Jonathon Luiten, Aljos̆a Os̆ep, Patrick Dendorfer, Philip Torr, Andreas Geiger, Laura Leal-Taixé, Bastian Leibe
2020 International Journal of Computer Vision  
To address this, we present a novel MOT evaluation metric, higher order tracking accuracy (HOTA), which explicitly balances the effect of performing accurate detection, association and localization into  ...  Multi-object tracking (MOT) has been notoriously difficult to evaluate. Previous metrics overemphasize the importance of either detection or association.  ...  For partial funding of this project, JL and BL would like to acknowledge the ERC Consolidator Grant Dee-ViSe (ERC-2017-COG-773161) and a Google Faculty Research Award.  ... 
doi:10.1007/s11263-020-01375-2 pmid:33642696 pmcid:PMC7881978 fatcat:jretenhya5hf7kv6f4h2h3jhlq

mvHOTA: A multi-view higher order tracking accuracy metric to measure spatial and temporal associations in multi-point detection [article]

Lalith Sharan, Halvar Kelm, Gabriele Romano, Matthias Karck, Raffaele De Simone, Sandy Engelhardt
2022 arXiv   pre-print
The main evaluation metric to benchmark MOT methods on datasets such as KITTI has recently become the higher order tracking accuracy (HOTA) metric, which is capable of providing a better description of  ...  In this work, we propose a multi-view higher order tracking metric (mvHOTA) to determine the accuracy of multi-point (multi-instance and multi-class) detection, while taking into account temporal and spatial  ...  Acknowledgements This work was supported in part by Informatics for Life funded by the Klaus Tschira Foundation and the German Research Foundation DFG Project 398787259, DE 2131/2-1 and EN 1197/2-1.).  ... 
arXiv:2206.09372v1 fatcat:mxijrzevv5bf7clnoyxk5ysc5i

MOTCOM: The Multi-Object Tracking Dataset Complexity Metric [article]

Malte Pedersen, Joakim Bruslund Haurum, Patrick Dendorfer, Thomas B. Moeslund
2022 arXiv   pre-print
There exists no comprehensive metric for describing the complexity of Multi-Object Tracking (MOT) sequences.  ...  of tracks.  ...  Depending on the metric, the ranking is in decreasing (HOTA) or increasing order (density, tracks, MOTCOM).  ... 
arXiv:2207.10031v1 fatcat:cw5oxnnn5vas5jougrhrwc3y54

BURST: A Benchmark for Unifying Object Recognition, Segmentation and Tracking in Video [article]

Ali Athar, Jonathon Luiten, Paul Voigtlaender, Tarasha Khurana, Achal Dave, Bastian Leibe, Deva Ramanan
2022 arXiv   pre-print
Multiple existing benchmarks involve tracking and segmenting objects in video e.g., Video Object Segmentation (VOS) and Multi-Object Tracking and Segmentation (MOTS), but there is little interaction between  ...  Additionally, we demonstrate several baselines for all tasks and show that approaches for one task can be applied to another with a quantifiable and explainable performance difference.  ...  We thank Jonas Schult, Christian Schmidt, Alexey Nekrasov, Sabarinath Mahadevan and Markus Knoche for helpful discussions.  ... 
arXiv:2209.12118v1 fatcat:ymfftq7ztvev5died224nbnfwq

AnimalTrack: A Large-scale Benchmark for Multi-Animal Tracking in the Wild [article]

Libo Zhang, Junyuan Gao, Zhen Xiao, Heng Fan
2022 arXiv   pre-print
Multi-animal tracking (MAT), a multi-object tracking (MOT) problem, is crucial for animal motion and behavior analysis and has many crucial applications such as biology, ecology, animal conservation and  ...  To address this problem, we introduce AnimalTrack, a large-scale benchmark for multi-animal tracking in the wild.  ...  Specifically, we employ the recently proposed higher order tracking accuracy (HOTA) from [31] , commonly used CLEAR metrics from [4] including multiple object tracking accuracy (MOTA), mostly tracked  ... 
arXiv:2205.00158v1 fatcat:a3qux6crazbvvlbo2urzgoz4ge

DeepFusionMOT: A 3D Multi-Object Tracking Framework Based on Camera-LiDAR Fusion with Deep Association [article]

Xiyang Wang, Chunyun Fu, Zhankun Li, Ying Lai, Jiawei He
2022 arXiv   pre-print
In the recent literature, on the one hand, many 3D multi-object tracking (MOT) works have focused on tracking accuracy and neglected computation speed, commonly by designing rather complex cost functions  ...  This association mechanism realizes tracking of an object in a 2D domain when the object is far away and only detected by the camera, and updating of the 2D trajectory with 3D information obtained when  ...  Evaluation Metrics:2D MOT Evaluation Metrics: CLEAR [38] is a commonly used method for evaluating 2D MOT performance, and it includes important evaluation metrics such as Multi-Object Tracking Accuracy  ... 
arXiv:2202.12100v1 fatcat:khfizcj2abdi5h7km4qfgndatu

DanceTrack: Multi-Object Tracking in Uniform Appearance and Diverse Motion [article]

Peize Sun, Jinkun Cao, Yi Jiang, Zehuan Yuan, Song Bai, Kris Kitani, Ping Luo
2022 arXiv   pre-print
A typical pipeline for multi-object tracking (MOT) is to use a detector for object localization, and following re-identification (re-ID) for object association.  ...  To this end, we propose a large-scale dataset for multi-human tracking, where humans have similar appearance, diverse motion and extreme articulation.  ...  We appreciate Xinshuo Weng, Yifu Zhang for valuable discussion and suggestions.  ... 
arXiv:2111.14690v3 fatcat:eccbcgkftjd5hjvrmd3gh2jeoi

PAE: Portable Appearance Extension for Multiple Object Detection and Tracking in Traffic Scenes

Ibrahim Soliman Mohamed, Lim Kim Chuan
2022 IEEE Access  
Multi-object tracking (MOT) is an important field in computer vision that provides a critical understanding of video analysis in various applications, such as vehicle tracking in intelligent transportation  ...  This study proposes a portable appearance extension (PAE) for single-stage object detection to jointly detect and extract appearance embeddings using a shared model.  ...  EVALUATION METRIC Our models are evaluated with HOTA (Higher Order Tracking Accuracy) [13] , [41] metric in the testing dataset.  ... 
doi:10.1109/access.2022.3160424 fatcat:dcftc5vpwjf7rcism6y3lvbsxm

How Trustworthy are Performance Evaluations for Basic Vision Tasks? [article]

Tran Thien Dat Nguyen, Hamid Rezatofighi, Ba-Ngu Vo, Ba-Tuong Vo, Silvio Savarese, Ian Reid
2022 arXiv   pre-print
This paper examines performance evaluation criteria for basic vision tasks involving sets of objects namely, object detection, instance-level segmentation and multi-object tracking.  ...  Intersection over Union (IoU) threshold, making their evaluations unreliable. More importantly, there is no means to verify whether we can trust the evaluations of a criterion.  ...  Metrics for Sets of Tracks For performance evaluation of multi-object tracking, the metrics for sets of shapes discussed earlier are not directly applicable because a track cannot be treated as a shape  ... 
arXiv:2008.03533v4 fatcat:j6r4okt4vvhxxogvsfqsypgn5e

Multi-Object Tracking and Segmentation with a Space-Time Memory Network [article]

Mehdi Miah, Guillaume-Alexandre Bilodeau, Nicolas Saunier
2021 arXiv   pre-print
We propose a method for multi-object tracking and segmentation that does not require fine-tuning or per benchmark hyper-parameter selection.  ...  We evaluated our tracker on KITTIMOTS and MOTSChallenge and show the benefit of our data association strategy with the HOTA metric. The project page is .  ...  Until recently, multi-object tracking and segmentation (MOTS) [37] were generally evaluated with measures heavily biased against the association step [20, 35] . Luiten et al.  ... 
arXiv:2110.11284v1 fatcat:yuj2ovj2djbrfo54kernxy7fdi

Tracking Every Thing in the Wild [article]

Siyuan Li, Martin Danelljan, Henghui Ding, Thomas E. Huang, Fisher Yu
2022 arXiv   pre-print
Current multi-category Multiple Object Tracking (MOT) metrics use class labels to group tracking results for per-class evaluation.  ...  We introduce a new metric, Track Every Thing Accuracy (TETA), breaking tracking measurement into three sub-factors: localization, association, and classification, allowing comprehensive benchmarking of  ...  Recently, Higher-Order Tracking Accuracy (HOTA) [30] was proposed to fairly balance both components by computing a separate score for each. Liu et al.  ... 
arXiv:2207.12978v1 fatcat:bqalvyjyofcjtdlghvegn7irmi

SimpleTrack: Rethinking and Improving the JDE Approach for Multi-Object Tracking [article]

Jiaxin Li and Yan Ding and Hualiang Wei
2022 arXiv   pre-print
Joint detection and embedding (JDE) based methods usually estimate bounding boxes and embedding features of objects with a single network in Multi-Object Tracking (MOT).  ...  To further improve the performance of data association, we develop a simple, effective tracker named SimpleTrack, which designs a bottom-up fusion method for Re-identity and proposes a new tracking strategy  ...  framework called SimpleTrack for data assocaition and multi-object tracking.  ... 
arXiv:2203.03985v1 fatcat:fw6eghrin5fbtdy652fooio2aq

CAMO-MOT: Combined Appearance-Motion Optimization for 3D Multi-Object Tracking with Camera-LiDAR Fusion [article]

Li Wang, Xinyu Zhang, Wenyuan Qin, Xiaoyu Li, Lei Yang, Zhiwei Li, Lei Zhu, Hong Wang, Jun Li, Huaping Liu
2022 arXiv   pre-print
As existing multi-object tracking methods only consider a single category, we also propose to build a multi-category loss to implement multi-object tracking in multi-category scenes.  ...  However, camera-based methods suffer in the case of occlusions and it can be challenging to accurately track the irregular motion of objects for LiDAR-based methods.  ...  ACKNOWLEDGMENTS We thank LetPub ( for linguistic assistance and pre-submission expert review.  ... 
arXiv:2209.02540v3 fatcat:ctjphkdglnfw3dnatpvgwne5ru

Local Metrics for Multi-Object Tracking [article]

Jack Valmadre, Alex Bewley, Jonathan Huang, Chen Sun, Cristian Sminchisescu, Cordelia Schmid
2021 arXiv   pre-print
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.  ...  We also thank Jonathan Luiten, Vivek Rathod and Zhichao Lu for valuable discussions and feedback.  ... 
arXiv:2104.02631v1 fatcat:kapmq4pfrfcylpk4x4ce5mqkbi
« Previous Showing results 1 — 15 out of 117 results