Filters








12,235 Hits in 2.8 sec

Masking Strategies for Image Manifolds [article]

Hamid Dadkhahi, Marco F. Duarte
2016 arXiv   pre-print
Our goal is for the manifold learned from masked images to resemble its full image counterpart as closely as possible.  ...  We consider the problem of selecting an optimal mask for an image manifold, i.e., choosing a subset of the pixels of the image that preserves the manifold's geometric structure present in the original  ...  ACKNOWLEDGEMENTS We thank Benjamin Marlin, Deepak Ganesan, and Addison Mayberry for helpful discussions, and for providing power consumption figures for the eyeglasses dataset.  ... 
arXiv:1606.04618v1 fatcat:bjahqbxd3bcnfcxsy2opvlos2i

Panoramic imaging system for mobile devices

Yingen Xiong, Xianglin Wang, Marius Tico, Chia-Kai Liang, Kari Pulli
2009 SIGGRAPH '09: Posters on - SIGGRAPH '09  
In the mask-based blending, we first create a mask for each source image. The mask has highest values at the image center and the lowest values on the borders.  ...  Our registration method for panorama application uses a coarse-to-fine strategy and a combination between image and feature based registration methods.  ...  In the mask-based blending, we first create a mask for each source image. The mask has highest values at the image center and the lowest values on the borders.  ... 
doi:10.1145/1599301.1599337 dblp:conf/siggraph/XiongWTLP09 fatcat:2shshzdxs5h7hdpfb4worwxddy

Performance assessment of a recent change detection method for homogeneous and heterogeneous images

Jorge Prendes, Marie Chabert, Frédéric Pascal, Alain Giros, Jean-Yves Tourneret
2015 Revue Francaise de Photogrammetrie et de Teledetection  
The estimated parameters are finally used to learn the relationships between the images of interest, via a manifold learning strategy.  ...  The main objective of this paper is to evaluate the performance of a change detection method based on this manifold learning strategy initially introduced in (Prendes et al., 2014a,b).  ...  Fig. 3 illustrates this manifold learning strategy for synthetic data, where the manifold is defined by the re- v(P ) = [T IR (P ), T R (P ), T G (P ), T B (P ), T SAR (P )] lationship T SAR (P ) = T  ... 
doi:10.52638/rfpt.2015.216 fatcat:5wmdj3527reujk5mrzbma4xqou

Joint recovery of under sampled signals on a manifold: Application to free breathing cardiac MRI

Sunrita Poddar, Sajan Goud Lingala, Mathews Jacob
2014 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We introduce novel algorithms for the joint recovery of an ensemble of signals that live on a smooth manifold from their under sampled measurements.  ...  of the underlying manifold structure.  ...  We conclude that for our reconstruction scheme, it is beneficial to sample signals lying on a manifold using incoherent sampling masks since it results in better recovery.  ... 
doi:10.1109/icassp.2014.6854938 dblp:conf/icassp/PoddarLJ14 fatcat:dy3bqqztz5bq5hjmd7htozkqqm

Image masking schemes for local manifold learning methods

Hamid Dadkhahi, Marco F. Duarte
2015 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We consider the problem of selecting a subset of the dimensions of an image manifold that best preserves the underlying local structure in the original data.  ...  We have previously shown that masks which preserve the data neighborhood graph are well suited to global manifold learning algorithms.  ...  MASKING STRATEGIES FOR LOCAL MANIFOLD LEARNING METHODS We propose a greedy algorithm for selection of an LLEaware masking pattern that attempts to preserve the weights w ij obtained from the optimization  ... 
doi:10.1109/icassp.2015.7179077 dblp:conf/icassp/DadkhahiD15 fatcat:2j3cu5rlbvat7pluxt3ucq5wh4

Evolving Image Compositions for Feature Representation Learning [article]

Paola Cascante-Bonilla, Arshdeep Sekhon, Yanjun Qi, Vicente Ordonez
2022 arXiv   pre-print
Although PatchMix can rely on random pairings and random grid-like patterns for mixing, we explore evolutionary search as a guiding strategy to jointly discover optimal grid-like patterns and image pairings  ...  Convolutional neural networks for visual recognition require large amounts of training samples and usually benefit from data augmentation.  ...  Acknowledgements The authors would like to thank the anonymous reviewers for their insightful comments. This work is supported by the National Science Foundation under awards No #2045773 and #2040961.  ... 
arXiv:2106.09011v2 fatcat:4twxldfjijgojlco36padd7tci

Optimization of Visual Information Presentation for Visual Prosthesis

Fei Guo, Yuan Yang, Yong Gao
2018 International Journal of Biomedical Imaging  
This paper proposes two image processing strategies based on a salient object detection technique.  ...  The results will provide effective solutions for the further development of visual prosthesis.  ...  The salient object detection model is based on the manifold ranking. The manifold ranking method proposed by Zhou et al. uses the intrinsic manifold structure of a data for graph labeling [18] .  ... 
doi:10.1155/2018/3198342 pmid:29731769 pmcid:PMC5872628 fatcat:neiub3rr6jevjincqnyrfcl7qm

Patch Alignment Manifold Matting

Xuelong Li, Kang Liu, Yongsheng Dong, Dacheng Tao
2018 IEEE Transactions on Neural Networks and Learning Systems  
In this paper, a manifold matting framework named Patch Alignment Manifold Matting is proposed for image matting.  ...  Finally, we apply some manifold learning methods in the framework, and obtain several image matting methods, such as named ISOMAP matting and its derived Cascade ISOMAP matting.  ...  Image mask α 1: Given an image, define a local neighborhood window w i for each pixel i; 2: Assume that the color space (RGB) can be approximated by manifold subspace in local patch; and utilize manifold  ... 
doi:10.1109/tnnls.2017.2727140 pmid:28783643 fatcat:mek5rmxlpvbfhck45oy6vwculm

Deep Manifold Learning for Dynamic MR Imaging [article]

Ziwen Ke, Zhuo-Xu Cui, Wenqi Huang, Jing Cheng, Sen Jia, Haifeng Wang, Xin Liu, Hairong Zheng, Leslie Ying, Yanjie Zhu, Dong Liang
2021 arXiv   pre-print
Specifically, the designed low-rank manifold provides a new technical route for applying low-rank priors in dynamic MR imaging.  ...  Methods: Cardiac MR image reconstruction is modeled as general compressed sensing (CS) based optimization on a low-rank tensor manifold.  ...  The Sensitivity to Different Undersampling Masks The proposed Manifold-Net has good reconstruction performance for different undersampling masks.  ... 
arXiv:2104.01102v1 fatcat:erywtclpzzfbpcl6nji4b3sbsq

DeepProjection: Rapid and structure-specific projections of tissue sheets embedded in 3D microscopy stacks using deep learning [article]

Daniel Haertter, Xiaolei Wang, Stephanie M Fogerson, Nitya Ramkumar, Janice M Crawford, Kenneth D Poss, Stefano Di Talia, Daniel P Kiehart, Christoph F Schmidt
2021 bioRxiv   pre-print
This algorithm rapidly and robustly extracts image content contained in curved manifolds from time-lapse recorded 3D image stacks by binary masking of background content, stack by stack.  ...  The masks calculated for a given movie, when predicted, e.g., on fluorescent cell boundaries on one channel, can subsequently be applied to project other fluorescent channels from the same manifold.  ...  Acknowledgements We thank David Carlson, Duke University, for input during the algorithm design process. D.H. thanks the German Academic Foundation for funding during this work.  ... 
doi:10.1101/2021.11.17.468809 fatcat:ucs33bhslbd5flvtuh6u4xdwm4

A Kernelized Manifold Mapping to Diminish the Effect of Adversarial Perturbations [article]

Saeid Asgari Taghanaki, Kumar Abhishek, Shekoofeh Azizi, Ghassan Hamarneh
2019 arXiv   pre-print
We evaluate the robustness of our method to different gradient (targeted and untargeted) and non-gradient based attacks and compare it to several non-gradient masking defense strategies.  ...  We test the proposed method on three publicly available image classification and segmentation datasets namely, MNIST, ISBI ISIC 2017 skin lesion segmentation, and NIH Chest X-Ray-14.  ...  The authors are grateful to the NVIDIA Corporation for donating a Titan X GPU used in this research.  ... 
arXiv:1903.01015v2 fatcat:myy2a2mquzddbpglws3ydzuwxy

Grassmann Registration Manifolds for Face Recognition [chapter]

Yui Man Lui, J. Ross Beveridge
2008 Lecture Notes in Computer Science  
Second, we embed the approximated tangent spaces on a Grassmann manifold and employ a chordal distance as the means for comparing subspaces.  ...  Motivated by image perturbation and the geometry of manifolds, we present a novel method combining these two elements.  ...  Bolme from Colorado State University for calculating the error bounds for the FERET data sets.  ... 
doi:10.1007/978-3-540-88688-4_4 fatcat:f5fgpd277fhubcxdldypzo6u6i

Mask Embedding in conditional GAN for Guided Synthesis of High Resolution Images [article]

Yinhao Ren, Zhe Zhu, Yingzhou Li, Joseph Lo
2019 arXiv   pre-print
To use semantic masks as guidance whilst providing realistic synthesized results with fine details, we propose to use mask embedding mechanism to allow for a more efficient initial feature projection in  ...  Semantic masks, such as sketches and label maps, are another intuitive and effective form of guidance in image synthesis.  ...  A mask input does not identify an image in the real image dataset, but instead relate to a cluster of real images. Hence masks gathered from dataset defines a partition of the real image manifolds.  ... 
arXiv:1907.01710v1 fatcat:6pbyazw4jfdovcixfedc6rkf4e

Post-DAE: Anatomically Plausible Segmentation via Post-Processing with Denoising Autoencoders [article]

Agostina J Larrazabal and César Martínez and Ben Glocker and Enzo Ferrante
2020 arXiv   pre-print
The proposed method is trained using unpaired segmentation mask, what makes it independent of intensity information and image modality.  ...  Then, given a segmentation mask obtained with an arbitrary method, we reconstruct its anatomically plausible version by projecting it onto the learnt manifold.  ...  ACKNOWLEDGMENTS We thank Alexandros Karargyris, Sema Candemir and Stefan Jaeger for sharing the segmentation masks used in the out-of-distribution experiments.  ... 
arXiv:2006.13791v1 fatcat:wzsnpqv2rzbzlbip6imfglei2y

A fully automated pipeline for brain structure segmentation in multiple sclerosis

Sandra González-Villá, Arnau Oliver, Yuankai Huo, Xavier Lladó, Bennett A. Landman
2020 NeuroImage: Clinical  
On the other hand, no significant volume differences were found when the proposed automated pipeline was used for segmentation, which demonstrates its robustness against variations in the lesion mask used  ...  segmented lesion masks.  ...  Segmentation result of m-JLF for the target image with (m) the manual and (n) the automatic lesion mask as input.  ... 
doi:10.1016/j.nicl.2020.102306 pmid:32585568 pmcid:PMC7322098 fatcat:neaxqicbrjfehlpmghthgx53ca
« Previous Showing results 1 — 15 out of 12,235 results