Semi-supervised On-Line Boosting for Robust Tracking

By Helmut Grabner, Christian Leistner, Horst Bischof. European Conference on Computer Vision, 2008

1,260 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/27d69a2d96600efb66fd907d8287ca3b6e734c59

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 Electrical and Electronic Engineering

Citations in the corpus

Understanding and Diagnosing Visual Tracking Systems (2015, methodology): “Thanks to advances made by machine learning researchers, many sophisticated techniques have been applied to visual tracking, including boosting [12, 13], multiple-instance learning [3], structured output SVM [14], Gaussian process regression [11], and deep learning [36, 34, 16].”
Tracking Revisited Using RGBD Camera: Unified Benchmark and Baselines (2013, methodology): “We also evaluated six state-of-theart RGB trackers: TLD[15], CT[30], MIL[3], semi-B[11], Struck[12] and VTD[16], since currently there is no other public available RGBD tracking algorithms for general objects.”
Object Tracking Benchmark (2015, methodology): “For discriminative models, the main issue has been improving the sample collection part to make the online-trained classiﬁer more robust [23, 5, 31, 26].”
In defense of color-based model-free tracking (2015, background): “Due to the success of discriminative models, a large variety of suitable learning methods has been explored for visual tracking, such as structured output SVMs [21], ranking SVMs [5], boosting [3, 18], kernel ridge regression [22, 23], and multipleinstance learning [4].”
Robust object tracking via sparsity-based collaborative model (2012, background): “the target object from the background [2, 10, 3, 23, 11, 4, 12].”
Sparse coding based visual tracking: Review and experimental comparison (2013, methodology): “To overcome the drifting problem, Grabner and Leistner [12] further proposed a semi-supervised boosting method to combine decisions of a given prior and an on-line classiﬁer.”
Learning a Deep Compact Image Representation for Visual Tracking (2013, background): “To address this problem, some trackers taking the semi-supervised learning approach have been proposed [12, 7].”
Robust tracking using local sparse appearance model and K-selection (2011, background): “In order to handle the drifting problem, semi-online boosting [10] was later proposed to incrementally update * This research is completed when the author is a research assistant in the UMDNJ-Robert Wood Johnson Medicial School (a) (b) (c) (d) (e) (f) Figure 1.”
Transferring Rich Feature Hierarchies for Robust Visual Tracking (2015, methodology): “Many advanced machine learning algorithms have been used, including boosting variants [10, 11], multiple-instance learning [2], structured output SVM [12], and Gaussian process regression [7].”
Robust Visual Tracking via Convolutional Networks Without Training (2015, background): “These results show that all the filers, soft shrinkage, and model update components play key roles in the proposed algorithm for robust visual tracking.”
Ensembles for feature selection: A review and future trends (2019, neutral): “Grabner et al. [78] presented a novel online boosting method which formulated the updated process in a semi-supervised fashion as combined decision of a given prior and an online classi(cid:12)er.”
Hough-based tracking of non-rigid objects (2011, positive): “[14] successfully apply semisupervised online boosting.”
Large Margin Object Tracking with Circulant Feature Maps (2017, methodology): “There are many classiﬁcation algorithms used in this framework, such as multiple instance learning [1], P-N learning [17], online boosting [11, 12], support vector machines (SVM) [13, 15, 22, 31] and so on.”
Correlation filters with limited boundaries (2014, methodology): “MOSSE KMOSSE MILTrack STRUCK OAB(1) SemiBoost FragTrack Our method [5] [13] [3] [11] [9] [10] [1] FaceOcc1 {1.”
Transfer Learning Based Visual Tracking with Gaussian Processes Regression (2014, methodology): “, a fixed prior classifier trained by the labeled samples from the first frame [7], a measurement matrix for compressive sensing [41], a fixed dictionary for histogram generation [45], or utilize the MUG estimation instead of the MAP estimation [19].”
Human Tracking Using Convolutional Neural Networks (2010, background): “In [ 27 ], a semi-supervised approach is proposed where labeled examples come from the first frame only, and subsequent training examples are left unlabeled.”
Struck: Structured output tracking with kernels (2011, positive): “A binary classifier generally ignores the unlabelled examples [10], while those based on semi-supervised learning use them in their update phase [11, 17].”