1,134 Hits in 4.8 sec

Light Field Reconstruction Using Convolutional Network on EPI and Extended Applications

Gaochang Wu, Yebin Liu, Lu Fang, Qionghai Dai, Tianyou Chai
2018 IEEE Transactions on Pattern Analysis and Machine Intelligence  
In this paper, a novel convolutional neural network (CNN)-based framework is developed for light field reconstruction from a sparse set of views.  ...  We evaluate our approach on several datasets, including synthetic scenes, real-world scenes and challenging microscope light field data.  ...  Then, in the "restoration" step, we apply a CNN to restore the angular detail of the EPI damaged by the undersampling.  ... 
doi:10.1109/tpami.2018.2845393 pmid:29994195 fatcat:xzuaadt2orb5lgopxs2ht3ejsm

A novel rendering approach for unstructured light field interpolation

Mandan Zhao, Xiangyang Hao
2017 Communications in Information and Systems  
What's more, a novel rendering approach is presented by combining proposed "blur-restoration-deblur" framework and depth information to handle large disparity and unstructured light field.  ...  In this paper, a novel framework is proposed for light field reconstruction from a sparse set of views. The light field is the unstructured, and has the large disparity.  ...  So we extend this approach for rendering novel views using multi-view input (e.g., multi-view stereo data) and a depth (disparity) map.  ... 
doi:10.4310/cis.2017.v17.n3.a3 fatcat:3minashktnb4jcxokrmrodt7ae

Depth Recovery Using an Adaptive Color-Guided Auto-Regressive Model [chapter]

Jingyu Yang, Xinchen Ye, Kun Li, Chunping Hou
2012 Lecture Notes in Computer Science  
This paper proposes an adaptive color-guided auto-regressive (AR) model for high quality depth recovery from low quality measurements captured by depth cameras.  ...  The AR predictor for each pixel is constructed according to both the local correlation in the initial depth map and the nonlocal similarity in the accompanied high quality color image.  ...  The rendered image suggests that our method reliably restores the geometric relationship from a set of low-resolution depth measurements.  ... 
doi:10.1007/978-3-642-33715-4_12 fatcat:tndpghkfqfflbcgk6lwkbigssi

Light Field Reconstruction Using Deep Convolutional Network on EPI

Gaochang Wu, Mandan Zhao, Liangyong Wang, Qionghai Dai, Tianyou Chai, Yebin Liu
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, we take advantage of the clear texture structure of the epipolar plane image (EPI) in the light field data and model the problem of light field reconstruction from a sparse set of views  ...  as a CNN-based angular detail restoration on EPI.  ...  We then apply a CNN to restore the angular detail of the EPI damaged by the undersampling. Finally, a non-blind deblur operation is used to restore the spatial detail suppressed by the EPI blur.  ... 
doi:10.1109/cvpr.2017.178 dblp:conf/cvpr/WuZWDCL17 fatcat:plubj7ayyvamdk5s5moqn5cq6y

Firearm Serial Number Restoration with Electron Backscatter Diffraction

Ryan M. White, Robert R. Keller
2015 Microscopy and Microanalysis  
Undersampling of the data in figure 2b provides a virtual pixel size increase, indicating that an unambiguous reconstruction can be performed with pixel sizes as large as 67.2 µm, decreasing acquisition  ...  It is yet unclear how this depth sensitivity compares to the traditional acid-etching-based methods of serial number reconstruction.  ...  Undersampling of the data in figure 2b provides a virtual pixel size increase, indicating that an unambiguous reconstruction can be performed with pixel sizes as large as 67.2 µm, decreasing acquisition  ... 
doi:10.1017/s1431927615007606 fatcat:ovyt5fa7uzcsljzxqxot3yfwja

Multi-layer Basis Pursuit for Compressed Sensing MR Image Reconstruction

Abdul Wahid, Jawad Shah, Adnan Umar Khan, Manzoor Ahmed, Hanif Razali
2020 IEEE Access  
Saiprasad Ravishankar for providing data for the CS-MRI analysis.  ...  of nonlinear mappings from undersampled MR images acquired in k -space.  ...  Sparse coding theory [41] works on the premise of first learning filters (weights/dictionaries) from given data and then finding the sparse representation maps from those dictionaries for representations  ... 
doi:10.1109/access.2020.3028877 fatcat:wdbm7fsngjgu5m5aaauudqsdze

Framelet pooling aided deep learning network: the method to process high dimensional medical data

Chang Min Hyun, Kang Cheol Kim, Hyun Cheol Cho, Jae Kyu Choi, Jin Keun Seo
2020 Machine Learning: Science and Technology  
One of the main difficulties is that there exists a computational cost problem in dealing with input data of large size matrices which represent medical images.  ...  Machine learning-based analysis of medical images often faces several hurdles, such as the lack of training data, the curse of the dimensionality problem, and generalization issues.  ...  Data availability statement Any data that support the findings of this study are included within the article.  ... 
doi:10.1088/2632-2153/ab592b fatcat:mep6loatdfbdxldpwaonjz2lhe

Restoration of firearm serial numbers with electron backscatter diffraction (EBSD)

Ryan M. White, Robert R. Keller
2015 Forensic Science International  
Undersampling the collected data by 5x (above center, 33.6 µm pixels) and 10x (above right, 67.2 µm pixels) still shows unambiguous reconstruction of the stamped letter.  ...  The image quality map of the cross section (above) shows that the deformation is detected via EBSD pattern quality mapping to a depth of about 520 µm beneath the bottom of the imprint.  ... 
doi:10.1016/j.forsciint.2015.02.003 pmid:25747326 fatcat:r74w2gjdarc27l5vlsrz4aqg4e

Depth Map Reconstruction with Local and Patch Manifold Regularized Deep Depth Priors

Abbas K. Ali, Peyman Adibi, Mohammad S. Ehsani
2021 IEEE Access  
In other words, most methods of ToF depth map restoration are also applicable to the Kinect depth map restoration [3] .  ...  In the current study, this equation shows the depth map, which is an inverse and inherently challenging ill-post problem in restoration x from y.  ... 
doi:10.1109/access.2021.3117140 fatcat:nsj7cfr46bamtdwbyvps5xjyyu

Color-Guided Depth Recovery From RGB-D Data Using an Adaptive Autoregressive Model

Jingyu Yang, Xinchen Ye, Kun Li, Chunping Hou, Yao Wang
2014 IEEE Transactions on Image Processing  
We observe and verify that the AR model tightly fits depth maps of generic scenes.  ...  This paper proposes an adaptive color-guided autoregressive (AR) model for high quality depth recovery from low quality measurements captured by depth cameras.  ...  Park for providing recovered depth maps in [11] for comparison, and thank anonymous reviewers for their comments which help to significantly improve the paper.  ... 
doi:10.1109/tip.2014.2329776 pmid:24951695 fatcat:5lfcex5awncrdeybablcwaa5uq

A novel image reconstruction method applied to deep Hubble Space Telescope Images [article]

Andrew Fruchter, Richard Hook
1997 arXiv   pre-print
We have developed a method for the linear reconstruction of an image from undersampled, dithered data, which has been used to create the distributed, combined Hubble Deep Field images -- the deepest optical  ...  The effect of undersampling on WF images is illustrated by the "Great Eye Chart in the Sky" in Figure 1 . Fortunately, much of the information lost to undersampling can be restored.  ...  Note that if the drop size is sufficiently small not all output pixels have data added to them from each input image.  ... 
arXiv:astro-ph/9708242v1 fatcat:rga5dvfkpnhitlneqjmoymmjzu

Deep multi-modal aggregation network for MR image reconstruction with auxiliary modality [article]

Chun-Mei Feng and Huazhu Fu and Tianfei Zhou and Yong Xu and Ling Shao and David Zhang
2022 arXiv   pre-print
Recently, many approaches have been developed to reconstruct full-sampled images from partially observed measurements to accelerate MR imaging.  ...  Magnetic resonance (MR) imaging produces detailed images of organs and tissues with better contrast, but it suffers from a long acquisition time, which makes the image quality vulnerable to say motion  ...  [15] were the first to propose a Convolutional Neural Network (CNN) to learn the mapping from an undersampled MR image to a reconstructed image.  ... 
arXiv:2110.08080v3 fatcat:aikmzbs525adhfj55fcnliv3ge

Swin Transformer for Fast MRI [article]

Jiahao Huang, Yingying Fang, Yinzhe Wu, Huanjun Wu, Zhifan Gao, Yang Li, Javier Del Ser, Jun Xia, Guang Yang
2022 arXiv   pre-print
A novel multi-channel loss was proposed by using the sensitivity maps, which was proved to reserve more textures and details.  ...  To accelerate the scanning process, methods by k-space undersampling and deep learning based reconstruction have been popularised.  ...  Method Classic Model-Based CS MRI Reconstruction To recover better spatial information with less artefacts from the undersampled k -space data, traditional CS-MRI methods usually consider solving the  ... 
arXiv:2201.03230v2 fatcat:6j3d5pjonzbs5ix3lua7pls7sq

The ModelCamera

V. Popescu, G. Bahmutov, E. Sacks, M. Mudure
2006 Graphical Models  
The operator selects views by checking the display for missing or undersampled surfaces and aiming the camera at them. A model is built from thousands of frames.  ...  The system registers the frames using the depth and color data, and integrates them into an evolving model that is displayed continually.  ...  Depth and possibly color data is acquired from several views. Depth is measured actively or is inferred from color. The data is registered in a common coordinate system.  ... 
doi:10.1016/j.gmod.2006.05.002 fatcat:mwcvibzcmvespodh7cmuwtlnky

Light Field Reconstruction with a Hybrid Sparse Regularization-Pseudo 4DCNN Framework

Li Fang, Wei Zhong, Long Ye, Li Ran, Qin Zhang
2020 IEEE Access  
[20] proposed a "blur-restoration-deblur" framework as learningbased angular detail restoration on 2D EPIs. Wang et al.  ...  as correspondence map for view reprojection.  ... 
doi:10.1109/access.2020.3023505 fatcat:l5d4uzk2g5dmzem34p2llmqkp4
« Previous Showing results 1 — 15 out of 1,134 results