Filters








1,036 Hits in 7.9 sec

An Anchor Patch Based Optimization Framework for Reducing Optical Flow Drift in Long Image Sequences [chapter]

Wenbin Li, Darren Cosker, Matthew Brown
2013 Lecture Notes in Computer Science  
In this paper, we propose an optimization framework that utilises a novel Anchor Patch algorithm which significantly reduces overall tracking errors given long sequences containing highly deformable objects  ...  Tracking through long image sequences is a fundamental research issue in computer vision.  ...  Conclusion In this paper, we have presented an optimization framework based on Anchor Patches for improving mesh or sparse point set tracking during long video image sequences.  ... 
doi:10.1007/978-3-642-37431-9_9 fatcat:fmrxxyzdcnaivfy723dqpnlob4

High-Detail 3D Capture and Non-sequential Alignment of Facial Performance

Martin Klaudiny, Adrian Hilton
2012 2012 Second International Conference on 3D Imaging, Modeling, Processing, Visualization & Transmission  
The mesh of an actor's face is tracked nonsequentially throughout a performance using multi-view image sequences.  ...  A robust patchbased frame-to-frame surface alignment algorithm combined with the optimal traversal significantly reduces drift compared to previous sequential techniques.  ...  ACKNOWLEDGMENT We would like to thank Alaleh Rashidnasab for being a test actor and T. Beeler et al. at ETH Zurich/Disney Research for releasing their dataset.  ... 
doi:10.1109/3dimpvt.2012.67 dblp:conf/3dim/KlaudinyH12 fatcat:2uibjxcbsfcdpgd5g45pogkoay

Deep incremental learning for efficient high-fidelity face tracking

Chenglei Wu, Takaaki Shiratori, Yaser Sheikh
2018 ACM Transactions on Graphics  
for the comparisons in Section 6.4, and Colin Lea and all reviewers for their constructive discussions and feedback.  ...  ACKNOWLEDGMENTS We would like to thank the actors for allowing us to use their data in Section 6.1, the authors of Beeler and colleagues [2011] and Fyffe and colleagues [2017] for providing their data  ...  Typical approaches in optical flow and scene flow computation compares rectangular patches in image domain. This assumption holds only if the target surface is frontal parallel.  ... 
doi:10.1145/3272127.3275101 fatcat:4bsyon3xuzfuzjn43pupxaye44

Continuous depth estimation for multi-view stereo

Yebin Liu, Xun Cao, Qionghai Dai, Wenli Xu
2009 2009 IEEE Conference on Computer Vision and Pattern Recognition  
Then, several depth candidates are generated based on a multiple starting scales (MSS) framework.  ...  From these candidates, refined depth maps for each view are synthesized according to path-based NCC (normalized cross correlation) metric.  ...  Figure 4 . 4 Patch-based NCC measurement based on continuous variational flow and surface patch.  ... 
doi:10.1109/cvprw.2009.5206712 fatcat:4yu34226lzemhdspqgbupclj64

Continuous depth estimation for multi-view stereo

Yebin Liu, Xun Cao, Qionghai Dai, Wenli Xu
2009 2009 IEEE Conference on Computer Vision and Pattern Recognition  
Then, several depth candidates are generated based on a multiple starting scales (MSS) framework.  ...  From these candidates, refined depth maps for each view are synthesized according to path-based NCC (normalized cross correlation) metric.  ...  Figure 4 . 4 Patch-based NCC measurement based on continuous variational flow and surface patch.  ... 
doi:10.1109/cvpr.2009.5206712 dblp:conf/cvpr/LiuCDX09 fatcat:lw6ajzy3c5hx3ok722tgptpmby

Deep Learning for Underwater Visual Odometry Estimation

Bernardo Teixeira, Hugo Silva, Anibal Matos, Eduardo Silva
2020 IEEE Access  
Additionally, an extension of current work is proposed, in the form of a visual-inertial sensor fusion network aimed at correcting visual odometry estimate drift.  ...  Robot visual-based navigation faces several additional difficulties in the underwater context, which severely hinder both its robustness and the possibility for persistent autonomy in underwater mobile  ...  [80] proposed to combine unsupervised stereo optical flow estimation and monocular disparity with classical model-based optimization for camera pose estimation.  ... 
doi:10.1109/access.2020.2978406 fatcat:zjjpiqgol5bclksbob6lnrf2lu

Transforming multiple visual surveys of a natural environment into time-lapses

Shane Griffith, Frank Dellaert, Cédric Pradalier
2019 The international journal of robotics research  
In comparison with another approach based on using iterative closest point (ICP) plus a homography, our framework produced more and better-quality alignments.  ...  As data association across year-long variation in appearance continues to represent a formidable challenge, we present success with a map-centric approach, which builds on 3D vision for visual data association  ...  optical flow.  ... 
doi:10.1177/0278364919881205 fatcat:h46ipg5varfibfncmqjslfk5e4

Online Tracking and Relocation Based on a New Rotation-Invariant Haar-like Statistical Descriptor in Endoscopic Examination

Haifan Gong, Limin Chen, Changhao Li, Jun Zeng, Xichen Tao, Yue Wang
2020 IEEE Access  
In this work, we construct an online tracking and relocation framework based on the concept of detection and tracking, which is dramatically adapted to the inherent characteristics of the gastrointestinal  ...  biopsy image.  ...  The tracking methods based on correlation filtering (KCF, CSR-DCF), multi-instance learning (MIL), and optical flow (Median Flow) make it difficult to achieve ideal results.  ... 
doi:10.1109/access.2020.2994440 fatcat:pma4lcugefcvdke3xiai4gvufy

Tracking using Numerous Anchor points [article]

Tanushri Chakravorty, Guillaume-Alexandre Bilodeau, Eric Granger
2017 arXiv   pre-print
In this paper, an online adaptive model-free tracker is proposed to track single objects in video sequences to deal with real-world tracking challenges like low-resolution, object deformation, occlusion  ...  These parameters are adapted to retain consistent features that vote for the object location and that deal with outliers for long-term tracking scenarios.  ...  Acknowledgements This work was supported in part by FRQ-NT team grant #167442 and by REPARTI (Regroupement pour l'étude des environnements partagés intelligents répartis) FRQ-NT strategic cluster.  ... 
arXiv:1702.02012v2 fatcat:iz2mo6znwjgojh5i5xrpf3zqcu

The Seventh Visual Object Tracking VOT2019 Challenge Results

Matej Kristan, Amanda Berg, Linyu Zheng, Litu Rout, Luc Van Gool, Luca Bertinetto, Martin Danelljan, Matteo Dunnhofer, Meng Ni, Min Young Kim, Ming Tang, Ming-Hsuan Yang (+169 others)
2019 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)  
The evaluation included the standard VOT and other popular methodologies for short-term tracking analysis as well as the standard VOT methodology for long-term tracking analysis.  ...  Two new challenges have been introduced: (iv) VOT-RGBT2019 challenge focused on short-term tracking in RGB and thermal imagery and (v) VOT-RGBD2019 challenge focused on long-term tracking in RGB and depth  ...  So that, by averaging the depth from the depth image in the same mask could determine the depth of the target.  ... 
doi:10.1109/iccvw.2019.00276 dblp:conf/iccvw/KristanBZRGBDDN19 fatcat:ogwxim7cgjanxiwq7dddqs66gy

Robust visual tracking using template anchors

Luka Cehovin, Ales Leonardis, Matej Kristan
2016 2016 IEEE Winter Conference on Applications of Computer Vision (WACV)  
Acknowledgments: This research was in part supported by ARRS projects L2-6765 and P2-0214.  ...  [23] proposed running an online discriminative tracker in parallel with motion prediction from a dense optical flow and NCC detection.  ...  an optimized native implementation would run in real-time on an average modern computer.  ... 
doi:10.1109/wacv.2016.7477570 dblp:conf/wacv/CehovinLK16 fatcat:gdc5d3q7nnaspicfb5lvxtpjce

T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos

Kai Kang, Hongsheng Li, Junjie Yan, Xingyu Zeng, Bin Yang, Tong Xiao, Cong Zhang, Zhe Wang, Ruohui Wang, Xiaogang Wang, Wanli Ouyang
2017 IEEE transactions on circuits and systems for video technology (Print)  
Despite their effectiveness on still images, those frameworks are not specifically designed for object detection from videos.  ...  still-image detection frameworks when they are applied to videos.  ...  Our proposed framework is based on the popular still-image object detection frameworks and adds important components specifically designed for videos.  ... 
doi:10.1109/tcsvt.2017.2736553 fatcat:zcj3z5nrqfhx7inpie57mwisd4

Vision and Learning for Deliberative Monocular Cluttered Flight [article]

Debadeepta Dey, Kumar Shaurya Shankar, Sam Zeng, Rupesh Mehta, M. Talha Agcayazi, Christopher Eriksen, Shreyansh Daftry, Martial Hebert, J. Andrew Bagnell
2014 arXiv   pre-print
In this work we present the first implementation of receding horizon control, which is widely used in ground vehicles, with monocular vision as the only sensing mode for autonomous UAV flight in dense  ...  Cameras provide a rich source of information while being passive, cheap and lightweight for small and medium Unmanned Aerial Vehicles (UAVs).  ...  dense optical flow [16] implementation in OpenCV to compute for every patch the average, minimum and maximum optical flow values which are used as feature descriptors of that patch. • Radon Transform  ... 
arXiv:1411.6326v1 fatcat:evbbzon54rb4xnatdlhjumpjyy

A Survey on Deep Learning for Localization and Mapping: Towards the Age of Spatial Machine Intelligence [article]

Changhao Chen, Bing Wang, Chris Xiaoxuan Lu, Niki Trigoni, Andrew Markham
2020 arXiv   pre-print
Instead of creating hand-designed algorithms through exploitation of physical models or geometric theories, deep learning based solutions provide an alternative to solve the problem in a data-driven way  ...  In this work, we provide a comprehensive survey, and propose a new taxonomy for localization and mapping using deep learning.  ...  Acknowledgments This work is supported by the EPSRC Project "ACE-OPS: From Autonomy to Cognitive assistance in Emergency OPerationS" (Grant Number: EP/S030832/1).  ... 
arXiv:2006.12567v2 fatcat:snb2byqamfcblauw5lzccb7umy

Learning Scale-Adaptive Tight Correlation Filter for Object Tracking

Shunli Zhang, Wei Lu, Weiwei Xing, Li Zhang
2018 IEEE Transactions on Cybernetics  
In this formulation, the correlation filter is set as the same size as the target, which can make full use of the relations of the adjacent image patches and effectively exclude the influence of the background  ...  First, we develop a tight correlation filter-based tracking framework from the signal detection perspective.  ...  In practice, the correlation filter can be learned based on the known image patch and its response, and applied for filtering in the testing image.  ... 
doi:10.1109/tcyb.2018.2868782 fatcat:k3hgqydynbbrfhbsiknak3ibcu
« Previous Showing results 1 — 15 out of 1,036 results