The Visual Object Tracking VOT2015 Challenge Results

By Matej Kristan, Ales Leonardis, Jiri Matas, Michael Felsberg, Roman P. Pflugfelder, L. Č. Zajc, Tomás Vojír, Gustav Häger, Alan Lukezic, A. Eldesokey and 92 more. 2015

901 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Evaluation & Benchmarks

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

In Evaluation & Benchmarks

Citations in the corpus

Learning feed-forward one-shot learners (2016, methodology): “We consider, however, that our implementation serves mostly as a proof-of-concept, using tracking as an interesting demonstration of one-shot-learning, and is orthogonal to many technical improvements found in the tracking literature [10].”
Learning Multi-domain Convolutional Neural Networks for Visual Tracking (2015, methodology): “For offline training of MDNet, we use 58 training sequences collected from VOT2013 [27], VOT2014 [26] and VOT2015 [25], excluding the videos included in OTB100.”
Fully-Convolutional Siamese Networks for Object Tracking (2016, methodology): “VOT-15 results.”
YouTube-BoundingBoxes: A Large High-Precision Human-Annotated Data Set for Object Detection in Video (2017, background): “• VOT [27] and MOT [30] are yearly visual object tracking challenges with associated data sets.”
One-Shot Video Object Segmentation (2016, methodology): “Evaluation as a tracker: Video object segmentation could also be evaluated as a Visual Object Tracking (VOT) [28] algorithm, by computing the bounding box around each of the segmentations.”
Discriminative Correlation Filter Tracker with Channel and Spatial Reliability (2016, methodology): “ated state-ofthe-art boundary-constraints formulations showed signiﬁ- cant beneﬁts of using our formulation. The CSR-DCF has state-of-the-art performance on standard benchmarks – OTB100 [44], VOT2015 [26] and VOT2016 [26] while running in real-time on a single CPU. Despite using simple features like”
Attentional Correlation Filter Network for Adaptive Visual Tracking (2017, methodology): “To train the attention network for evaluating on the CVPR2013 and TPAMI2015 datasets, the VOT2014 [22] and VOT2015 [21] datasets were used.”
MCDNN: An Approximation-Based Execution Framework for Deep Stream Processing Under Resource Constraints (2016, background): “Although traditionally detection and tracking have been performed with relatively lightweight algorithms, state-of-the-art variants of these are switching to Deep Neural Networks (DNNs) [29, 33].”
The Visual Object Tracking VOT2016 Challenge Results (2016, methodology): “In VOT2015 [18], the methodology was extended into a fully automated sequence selection with the selection process focusing on challenging sequences.”
A Twofold Siamese Network for Real-Time Object Tracking (2018, methodology): “VOT2015 benchmark: VOT2015 is one of the most popular object tracking benchmarks.”
ECO: Efficient Convolution Operators for Tracking (2016, methodology): “The EAO measure, computed as the average EAO over typical sequence lengths (grey region), is displayed in the legend (see [24] for details).”
Vision Meets Drones: A Challenge (2018, neutral): “[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, methodology): “Note that VOT15 [9] is sampled from existing datasets like OTB100 [5] and ALOV300 [17], resulting in overlapping sequences ( e.g. basketball, car, singer, etc... ).”
Learning Background-Aware Correlation Filters for Visual Tracking (2017, methodology): “We extensively evaluate our tracker on four standard datasets, including OTB50 [36], OTB100 [37], TempleColor128 (TC128) [23], and VOT2015 [18], comparing with 24 state-of-the-art methods, such as TLD [15], Struck [12], CFLB [17], KCF [13], DSST [7], SAMF [22], MEEM [38], DAT [29], LCT [27], HCF [26”
Online Multi-Object Tracking with Dual Matching Attention Networks (2018, methodology): “The ECO tracker achieves the state-of-the-art performance on visual tracking benchmarks [25,52,38,33] and its fast variant ECO-HC based on hand-crafted features (HOG [11] and Color Names [49]) operates at 60 frames per second (FPS) on a single CPU, which is suitable for the online MOT task.”
Meta-Tracker: Fast and Robust Online Adaptation for Visual Object Trackers (2018, background): “The expected average overlap is an estimator of the average overlap a tracker is expected to attain on a large collection of short-term sequences(The more details in [28]).”
Deeper and Wider Siamese Networks for Real-Time Visual Tracking (2019, methodology): “The listed methods, such as EBT[41], LDP[24], nSAMF[22], TCNN[25], MLDF[36], CFWCR[36] and CFCF[11] are compared in VOT challenges [9, 16, 17].”