A survey of appearance models in visual object tracking

By Xi Li, Weiming Hu, Chunhua Shen, Zhongfei Zhang, Anthony R. Dick, Anton van den Hengel. ACM Transactions on Intelligent Systems and Technology, 2013

771 citations recorded by the source corpus.

Visual object tracking is a significant computer vision task which can be applied to many domains, such as visual surveillance, human computer interaction, and video compression. Despite extensive research on this topic, it still suffers from difficulties in handling complex object appearance changes caused by factors such as illumination variation, partial occlusion, shape deformation, and camera motion. Therefore, effective modeling of the 2D appearance of tracked objects is a key issue for the success of a visual tracker. In the literature, researchers have proposed a variety of 2D appearance models. To help readers swiftly learn the recent advances in 2D appearance models for visual object tracking, we contribute this survey, which provides a detailed review of the existing 2D appearance models. In particular, this survey takes a module-based architecture that enables readers to easily grasp the key points of visual object tracking. In this survey, we first decompose the problem of appearance modeling into two different processing stages: visual representation and statistical modeling. Then, different 2D appearance models are categorized and discussed with respect to their composition modules. Finally, we address several issues of interest as well as the remaining challenges for future research on this topic. The contributions of this survey are fourfold. First, we review the literature of visual representations according to their feature-construction mechanisms (i.e., local and global). Second, the existing statistical modeling schemes for tracking-by-detection are reviewed according to their model-construction mechanisms: generative, discriminative, and hybrid generative-discriminative. Third, each type of visual representations or statistical modeling techniques is analyzed and discussed from a theoretical or practical viewpoint. Fourth, the existing benchmark resources (e.g., source codes and video datasets) are examined in this survey.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/506f628df107dc62abc4acbe403fe9dfbe06684a

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

MEEM: Robust Tracking via Multiple Experts Using Entropy Minimization (2014, background): “For more comprehensive literature reviews, readers are directed to [18,28].”
The Visual Object Tracking VOT2015 Challenge Results (2015, background): “Visual tracking is diverse research area that has attracted significant attention over the last fifteen years [20, 48, 18, 27, 49, 79, 43].”
The Visual Object Tracking VOT2016 Challenge Results (2016, background): “The signiﬁcant activity in the ﬁeld over last two decades is reﬂected in the abundance of review papers [1–9].”
Empowering Things With Intelligence: A Survey of the Progress, Challenges, and Opportunities in Artificial Intelligence of Things (2020, background): “background context information to train an online discriminative classifier [71].”
Distributed Deep Neural Networks Over the Cloud, the Edge and End Devices (2017, background): “5In practical object tracking systems, this synchronization step is typically automated [19].”
Online Multi-Expert Learning for Visual Tracking (2020, neutral): “There are various kinds of visual tracking algorithms, and a brief review can be found in [1], [23]–[25].”
Online Object Tracking: A Benchmark (2013, background): “Object representation is one of major components in any visual tracker and numerous schemes have been presented [35].”
Visual Tracking With Weighted Adaptive Local Sparse Appearance Model via Spatio-Temporal Context Learning (2018, background): “V ISUAL tracking is an important research topic in computer vision with wide applications in various fields, such as self-driving cars, security and surveillance systems, intelligent transportation vision based controls [1].”
Object Tracking Benchmark (2015, background): “Object representation is one of major components in any visual tracker and numerous schemes have been presented [35].”
Robust Visual Tracking via Convolutional Networks Without Training (2015, background): “I. INTRODUCTION V ISUAL tracking is a fundamental problem in computervision with a wide range of applications.”
A Novel Performance Evaluation Methodology for Single-Target Trackers (2015, methodology): “Index Terms—Performance analysis, single-target tracking, model-free tracking, tracker evaluation methodology, tracker evaluation datasets, tracker evaluation system F”
Towards More Flexible and Accurate Object Tracking with Natural Language: Algorithms and Benchmark (2021, background): “For more related works on standard visual tracking, please check the following survey papers [33, 35, 41, 48, 65].”
Deeper and Wider Siamese Networks for Real-Time Visual Tracking (2019, background): “Tracking at real-time speeds plays a vital role in numerous vision applications, such as surveillance, robotics, and human-computer interaction [10, 21, 30, 43].”
The Visual Object Tracking VOT2014 Challenge Results (2014, neutral): “The significant activity in the field is also reflected in the abundance of review papers [23,43,22,29,44,65,40] summarizing the advances published in conferences and journals over the last fifteen years.”
Visual Tracking via Adaptive Spatially-Regularized Correlation Filters (2019, background): “Visual tracking [36, 25, 24] is a fundamental computer vision problem and has many realistic applications including video surveillance, behavior analysis, to name a few.”
TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild (2018, background): “Note that the maximum sequence length is 30, thus t ∈ [1, 30].”
SeqTrack: Sequence to Sequence Learning for Visual Object Tracking (2023, background): “Prior knowledge, such as window penalty [2,27] and on-line update [12, 28, 53], has been widely incorporated into existing tracking models and proved effective [7,10,43,55].”