The Visual Object Tracking VOT2014 Challenge Results

By Matej Kristan, Juan E. Sala Matas, Ales Leonardis, Jiri Matas, Luka Cehovin, Georg Nebehay, Tomás Vojír, Gustavo Javier Fernandez, Alan Lukezic, A. Dimitriev and 47 more. ECCV Workshops, 2014

512 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/3c74b636c0f74c1a0cbbd6e165c2760264044971

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Attentional Correlation Filter Network for Adaptive Visual Tracking (2017, methodology): “For the evaluation on the VOT2014 dataset, we trained the attention network by 39 sequences of the CVPR2013 dataset after eliminating overlapping scenes.”
Understanding and Diagnosing Visual Tracking Systems (2015, background): “To address this fairness concern, several relatively large benchmarks [39, 18] and evaluation metrics [6] have been proposed recently.”
Learning to Track at 100 FPS with Deep Regression Networks (2016, methodology): “The VOT 2014 Tracking Challenge consists of two types of experiments.”
Need for Speed: A Benchmark for Higher Frame Rate Object Tracking (2017, background): “VOT14 and VOT15: VOT14 [18] and VOT15 [19] consist of 25 and 30 challenging videos, respectively.”
In defense of color-based model-free tracking (2015, methodology): “Figure 4: Ranking results on the benchmark datasets of the (a) VOT14 and (b) VOT13 challenges.”
Discriminative Scale Space Tracking (2016, methodology): “Here we present the results on the VOT 2014 dataset [8].”
Learning Spatially Regularized Correlation Filters for Visual Tracking (2015, methodology): “Table 3 shows the ﬁnal ranking scores over all the videos in VOT2014.”
Struck: Structured Output Tracking with Kernels (2016, unknown): “It also forms an essential part of higher-level vision tasks such as scene understanding and action recognition.”
Real-Time MDNet (2018, methodology): “For this experiments, we pretrain our network using VOT-OTB dataset, which consist of 58 videos collected from VOT2013 [33], VOT2014 [34] and VOT2015 [35] excluding the videos in OTB2015.”
The Visual Object Tracking VOT2015 Challenge Results (2015, methodology): “The VOT2015 was a third attempt toward this, following the very successful VOT2013 and VOT2014.”
LaSOT: A High-Quality Benchmark for Large-Scale Single Object Tracking (2018, unknown): “Summaries of existing tracking benchmarks with highquality dense (per frame) annotations, including OTB-2013 [52], OTB-2015 [53], TC-128 [35], NUS-PRO [28], UAV123 [39], UAV20L [39], VOT-2014 [26], VOT-2017 [27] and LaSOT.”
Modeling and Propagating CNNs in a Tree Structure for Visual Tracking (2016, background): “• The proposed algorithm presents outstanding accuracy in the standard tracking benchmarks [37, 36, 19] and outperforms the state-of-the-art methods by large margins.”
Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning (2017, methodology): “In order to pre-train ADNet, we used 360 training videos from VOT2013 [19], VOT2014 [20], VOT2015 [18], and ALOV300 [29], in which videos overlapping with OTB50 and OTB-100 were excluded.”
Vision Meets Drones: A Challenge (2018, unknown): “[24], [38], [39] organize a VOT competition from 2013 to 2017 by presenting new datasets and evaluation strategies for tracking evaluation.”
TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild (2018, background): “Most trackers submitted to the annual challenge VOT17 [11] use deep features, while they were nonexistent in earlier versions VOT13 [7] and VOT14 [8].”
MOTChallenge 2015: Towards a Benchmark for Multi-Target Tracking (2015, background): “ralized benchmarks for numerous tasks including object detection [19], pedestrian detection [17], 3D reconstruction [45], optical ﬂow [9], [22], visual odometry [22], singleobject short-term tracking [28], and stereo estimation [22], [43]. Despite potential pitfalls of such benchmarks (e.g. [48]), t”