Learning a Deep Compact Image Representation for Visual Tracking

By Naiyan Wang, Dit-Yan Yeung. 2013

974 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

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

Rethinking the Inception Architecture for Computer Vision (2015, background): “…[16] winning entry by Krizhevsky et al [9], their network “AlexNet” has been successfully applied to a larger variety of computer vision tasks, for example to object-detection [5], segmentation [12], human pose estimation [22], video classiﬁcation [8], object tracking [23], and superresolution [3].”
Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning (2016, methodology): “Since the 2012 ImageNet competition [11] winning entry by Krizhevsky et al [8], their network “AlexNet” has been successfully applied to a larger variety of computer vision tasks, for example to object-detection [4], segmentation [10], human pose estimation [17], video classification [7], object tra”
Collaborative Deep Learning for Recommender Systems (2014, background): “On the other hand, deep learning models recently show great potential for learning effective representations and deliver state-of-the-art performance in computer vision [38] and natural language processing [15, 26] applications.”
Transferring Rich Feature Hierarchies for Robust Visual Tracking (2015, methodology): “These trackers include Struck [12], SCM [38], TLD [17], ASLA [16], DLT [30] and TGPR [7].”
Hedged Deep Tracking (2016, methodology): “Comparison with state-of-the-art trackers We compare our algorithm with 8 recent state-of-theart trackers: DLT [30], CNN-SVM [18], KCF [17], MEEM [36], Struck [14], CXT [8], TLD [20], and SCM [41].”
Online Multi-Expert Learning for Visual Tracking (2020, methodology): “In [40], the 80M tiny image dataset [41] has been used to pre-train a multi-layer autoencoder network.”
Unsupervised Deep Tracking (2019, methodology): “To the best of our knowledge, the only related but different approach is the auto-encoder based method [51].”
DeepLOB: Deep Convolutional Neural Networks for Limit Order Books (2018, methodology): “CNNs have been successfully applied to various application domains, for example, object tracking [31], object-detection [32] and segmentation [33].”
Universally Slimmable Networks and Improved Training Techniques (2019, background): “The ability to run neural network models within latency budget is of paramount importance for applications on mobile phones, augmented reality glasses, self-driving cars, security cameras and many others [2, 3, 4, 5, 6, 7, 8, 9].”
LaSOT: A High-quality Large-scale Single Object Tracking Benchmark (2020, background): “Inspired by the success of deep learning, many deep trackers [26, 68, 74, 79, 81] have been proposed and exhibit state-of-the-art performance.”
Hierarchical Convolutional Features for Visual Tracking (2015, methodology): “Wang and Yeung [30] propose a deep learning tracker (DLT) using a multi-layer autoencoder network.”
Convolutional Features for Correlation Filter Based Visual Tracking (2015, background): “The authors of [38] propose a compact deep feature based tracking framework that learns generic features by employing a stacked denoising auto-encoder.”
Robust Visual Tracking via Convolutional Networks Without Training (2015, methodology): “To validate the effectiveness of key components of CNT, we propose three variants of CNT: one utilizes random patch filters to replace the filters learned by k-means algorithm, one does not involve the soft shrinkage process, and another one does not employ the model update scheme (4).”
Multi-manifold deep metric learning for image set classification (2015, background): “…[20], deep convolutional neural networks [40], and deep belief network [12], and some of them have been successfully employed in various vision applications such as image classiﬁcation [17], object detection [30], action recognition [20], face veriﬁcation [31], and visual tracking [33].”
Learning to Track at 100 FPS with Deep Regression Networks (2016, background): “Second, most trackers evaluate a finite number of samples and choose the highest scoring one as the tracking output [26,27,37,30,39,24,33].”
DeepTrack: Learning Discriminative Feature Representations by Convolutional Neural Networks for Visual Tracking (2014, methodology): “Using offline trained CNN for tracking has been proposed in the past [8], [17].”