The Visual Object Tracking VOT2016 Challenge Results

Kristan, Matej; Leonardis, Aleš; Matas, Jiřı́; Felsberg, Michael; Pflugfelder, Roman; Čehovin, Luka; Vojíř, Tomáš; Hager, G; Lukežič, Alan; Fernández, Gustavo J.; Gupta, Abhinav; Petrosino, Alfredo; Memarmoghadam, Alireza; García-Martín, Álvaro; Montero, Andrés Solís; Vedaldi, Andrea; Robinson, Andreas; Andy, J.; Varfolomieiev, Anton; Alatan, A. Aydın; Erdem, Aykut; Ghanem, Bernard; Liu, Bin; Han, Bohyung; Martínez, Brais; Chang, Chang-Ming; Xu, Changsheng; Sun, Chong; Kim, Daijin; Chen, Dapeng; Du, Dawei; Mishra, Deepak; Yeung, Dit–Yan; Gundogdu, Erhan; Erdem, Erkut; Khan, Fahad Shahbaz; Porikli, Fatih; Zhao, Fei; Bunyak, Filiz; Battistone, Francesco; Zhu, Gao; Roffo, Giorgio; Subrahmanyam, Gorthi R. K. Sai; Bastos, Guilherme Sousa; Seetharaman, Guna; Medeiros, Henry; Li, Hongdong; Qi, Honggang; Bischof, Horst; Possegger, Horst; Lu, Huchuan; Lee, Hyemin; Nam, Hyeonseob; Chang, Hyung Jin; Drummond, Isabela; Valmadre, Jack; Jeong, Jae-chan; Cho, Jae-il; Lee, Jae-Yeong; Zhu, Jianke; Feng, Jiayi; Gao, Jin; Choi, Jin Young; Xiao, Jingjing; Kim, Ji Wan; Jeong, Jiyeoup; Henriques, João F.; Lang, Jochen; Choi, Jongwon; Martínez, José María Álvarez; Xing, Junliang; Gao, Junyu; Palaniappan, Kannappan; Lebeda, Karel; Gao, Ke; Mikolajczyk, Krystian; Qin, Lei; Wang, Lijun; Wen, Longyin; Bertinetto, Luca; Rapuru, Madan Kumar; Poostchi, Mahdieh; Maresca, Mario Edoardo; Danelljan, Martin; Mueller, Matthias; Zhang, Mengdan; Arens, Michael; Valstar, Michel; Tang, Ming; Baek, Mooyeol; Khan, Muhammad Haris; Wang, Naiyan; Fan, Nana; Al-Shakarji, Noor M.; Mikšík, Ondřej; Akin, Osman; Moallem, Payman; Senna, Pedro; Torr, Philip H. S.; Yuen, Pong C.; Huang, Qingming; Martín-Nieto, Rafael; Pelapur, Rengarajan; Bowden, Richard; Laganière, Robert; Stolkin, Rustam; Walsh, Ryan R.; Krah, Sebastian B.; Li, Shengkun; Zhang, Shengping; Yao, Shizeng; Hadfield, Simon; Melzi, Simone; Lyu, Siwei; Li, Siyi; Becker, Stefan; Golodetz, Stuart; Kakanuru, Sumithra; Choi, Sunglok; Hu, Tao; Mauthner, Thomas; Zhang, Tianzhu; Pridmore, Tony; Santopietro, Vincenzo; Hu, Weiming; Li, Wenbo; Hübner, Wolfgang; Lan, Xiangyuan; Wang, Xiaomeng; Li, Xin; Li, Yang; Demiris, Yiannis; Wang, Yifan; Qi, Yuankai; Yuan, Zejian; Cai, Zexiong; Xu, Zhan; He, Zhenyu; Chi, Zhizhen

doi:10.1007/978-3-319-48881-3_54

Cited by 501 publications

(685 citation statements)

References 90 publications

(6 reference statements)

Supporting

Mentioning

678

Contrasting

Unclassified

Order By: Relevance

“…To compare the performance of the many different approaches, a vast amount of evaluation datasets and schemes are available. They include large detection datasets with multiple object categories, such as PASCAL VOC [4], smaller, more specific detection datasets with a single category, such as cars [6], and sequences with multiple frames that are commonly used to evaluate trackers such as VOT2016 [7], OTB-2015 [20], or MOT16 [10]. Although very different in their nature, all of the benchmarks use axis-aligned or oriented boxes as ground truth and estimate the accuracy with the Intersection over Union (IoU) criterion.…”

Section: Introductionmentioning

confidence: 99%

“…For example, the (a) bag from VOT2016 [7] (b) blackswan from DAVIS [12] (c) boat from DAVIS [12] Figure 1: In image (a), both oriented boxes have an identical IoU with the ground truth segmentation. Nevertheless, their common IoU is only 0.71.…”

Section: Introductionmentioning

confidence: 99%

“…VOT2016 Benchmark [7] generates plausible oriented boxes from densely segmented objects and the COCO 2014 Detection challenge [9] uses axis-aligned bounding boxes of the segmentations to simplify the evaluation protocol. Hence, approaches may have a relatively low IoU with the ground truth, although their IoU with the actual object segmentation is the same (or even better) than that of the ground truth box (see Fig.…”

Section: Introductionmentioning

confidence: 99%

“…The proposed evaluation removes the bias introduced by restricting the ground truth to boxes for densely segmented data (such as COCO 2014 Detection Challenge [9] or VOT2016 [7]). …”

Section: Introductionmentioning

confidence: 99%

“…The proposed measure and evaluation scheme is evaluated on a handful of state-of-theart trackers for the DAVIS [12] and VOT2016 [7] datasets and will be made available to the community.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Pattern Recognition

Böttger

Follmann

Fauser

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The accuracy of object detectors and trackers is most commonly evaluated by the Intersection over Union (IoU) criterion. To date, most approaches are restricted to axis-aligned or oriented boxes and, as a consequence, many datasets are only labeled with boxes. Nevertheless, axis-aligned or oriented boxes cannot accurately capture an object's shape. To address this, a number of densely segmented datasets has started to emerge in both the object detection and the object tracking communities. However, evaluating the accuracy of object detectors and trackers that are restricted to boxes on densely segmented data is not straightforward. To close this gap, we introduce the relative Intersection over Union (rIoU) accuracy measure. The measure normalizes the IoU with the optimal box for the segmentation to generate an accuracy measure that ranges between 0 and 1 and allows a more precise measurement of accuracies. Furthermore, it enables an efficient and easy way to understand scenes and the strengths and weaknesses of an object detection or tracking approach. We display how the new measure can be efficiently calculated and present an easy-to-use evaluation framework. The framework is tested on the DAVIS and the VOT2016 segmentations and has been made available to the community.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…The proposed evaluation removes the bias introduced by restricting the ground truth to boxes for densely segmented data (such as COCO 2014 Detection Challenge [9] or VOT2016 [7]). …”

Section: Introductionmentioning

confidence: 99%

“…The proposed measure and evaluation scheme is evaluated on a handful of state-of-theart trackers for the DAVIS [12] and VOT2016 [7] datasets and will be made available to the community.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Pattern Recognition

Böttger

Follmann

Fauser

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

Unveiling the Power of Deep Tracking

Bhat

Johnander

Danelljan

et al. 2018

Computer Vision – ECCV 2018

Self Cite

398

243

View full text Add to dashboard Cite

In the field of generic object tracking numerous attempts have been made to exploit deep features. Despite all expectations, deep trackers are yet to reach an outstanding level of performance compared to methods solely based on handcrafted features. In this paper, we investigate this key issue and propose an approach to unlock the true potential of deep features for tracking. We systematically study the characteristics of both deep and shallow features, and their relation to tracking accuracy and robustness. We identify the limited data and low spatial resolution as the main challenges, and propose strategies to counter these issues when integrating deep features for tracking. Furthermore, we propose a novel adaptive fusion approach that leverages the complementary properties of deep and shallow features to improve both robustness and accuracy. Extensive experiments are performed on four challenging datasets. On VOT2017, our approach significantly outperforms the top performing tracker from the challenge with a relative gain of 17% in EAO.

show abstract

Meta-tracker: Fast and Robust Online Adaptation for Visual Object Trackers

Park

Berg

2018

Computer Vision – ECCV 2018

169

135

View full text Add to dashboard Cite

This paper improves state-of-the-art visual object trackers that use online adaptation. Our core contribution is an offline metalearning-based method to adjust the initial deep networks used in online adaptation-based tracking. The meta learning is driven by the goal of deep networks that can quickly be adapted to robustly model a particular target in future frames. Ideally the resulting models focus on features that are useful for future frames, and avoid overfitting to background clutter, small parts of the target, or noise. By enforcing a small number of update iterations during meta-learning, the resulting networks train significantly faster. We demonstrate this approach on top of the high performance tracking approaches: tracking-by-detection based MDNet [1] and the correlation based CREST [2]. Experimental results on standard benchmarks, OTB2015 [3] and VOT2016 [4], show that our meta-learned versions of both trackers improve speed, accuracy, and robustness.

show abstract

The Visual Object Tracking VOT2016 Challenge Results

Cited by 501 publications

References 90 publications

Pattern Recognition

Pattern Recognition

Unveiling the Power of Deep Tracking

Meta-tracker: Fast and Robust Online Adaptation for Visual Object Trackers

Contact Info

Product

Resources

About