14 Hits in 5.0 sec

Unrolling the Shutter: CNN to Correct Motion Distortions

Vijay Rengarajan, Yogesh Balaji, A. N. Rajagopalan
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Row-wise exposure delay present in CMOS cameras is responsible for skew and curvature distortions known as the rolling shutter (RS) effect while imaging under camera motion.  ...  We propose a convolutional neural network (CNN) architecture that automatically learns essential scene features from a single RS image to estimate the row-wise camera motion and undo RS distortions back  ...  In our method, we correct the distorted image geometrically using the motion estimate from the CNN.  ... 
doi:10.1109/cvpr.2017.252 dblp:conf/cvpr/RengarajanBR17 fatcat:vgar6xhgzzfrlfifntbr2pppcu

SUNet: Symmetric Undistortion Network for Rolling Shutter Correction [article]

Bin Fan and Yuchao Dai and Mingyi He
2021 arXiv   pre-print
In this paper, we present a novel deep network to solve the generic rolling shutter correction problem with two consecutive frames.  ...  The vast majority of modern consumer-grade cameras employ a rolling shutter mechanism, leading to image distortions if the camera moves during image acquisition.  ...  Note that the deep shutter unrolling network (DSUN) [20] is most relevant to our approach.  ... 
arXiv:2108.04775v1 fatcat:ltmgkgxmbvfkvljcmc4mjy5hjm

Context-Aware Video Reconstruction for Rolling Shutter Cameras [article]

Bin Fan, Yuchao Dai, Zhiyuan Zhang, Qi Liu, Mingyi He
2022 arXiv   pre-print
With the ubiquity of rolling shutter (RS) cameras, it is becoming increasingly attractive to recover the latent global shutter (GS) video from two consecutive RS frames, which also places a higher demand  ...  Specifically, we first estimate the bilateral motion field so that the pixels of the two RS frames are warped to a common GS frame accordingly.  ...  Rolling shutter correction advocates the mitigation or elimination of RS distortion, i.e. recovering the latent GS image, from a single frame [22, 43, 44, 64] or multiple frames [1, 15, 24, 47, 54,  ... 
arXiv:2205.12912v1 fatcat:5oihoxsgqjgrtbusaq6izad6c4

2021 Index IEEE Transactions on Pattern Analysis and Machine Intelligence Vol. 43

2022 IEEE Transactions on Pattern Analysis and Machine Intelligence  
Departments and other items may also be covered if they have been judged to have archival value. The Author Index contains the primary entry for each item, listed under the first author's name.  ...  -that appeared in this periodical during 2021, and items from previous years that were commented upon or corrected in 2021.  ...  ., +, TPAMI April 2021 1239-1253 Cascade R-CNN: High Quality Object Detection and Instance Segmentation. Correlation Correction to "Nonlinear Regression via Deep Negative Correlation Learning".  ... 
doi:10.1109/tpami.2021.3126216 fatcat:h6bdbf2tdngefjgj76cudpoyia

Unsupervised Domain-Specific Deblurring using Scale-Specific Attention [article]

Praveen Kandula, Rajagopalan. A. N
2021 arXiv   pre-print
For network training at a specific scale, SAAM attends to lower scale features as a function of the current scale.  ...  To address the above issues, we propose unsupervised domain-specific deblurring using a scale-adaptive attention module (SAAM).  ...  Unrolling the 369, 2018. 2 shutter: Cnn to correct motion distortions. In Proceedings of [20] A. Mittal, A. K. Moorthy, and A. C. Bovik.  ... 
arXiv:2112.06175v1 fatcat:rzz2t7dkonbaxp3ylx4li3alpm

GeoRefine: Self-Supervised Online Depth Refinement for Accurate Dense Mapping [article]

Pan Ji, Qingan Yan, Yuxin Ma, Yi Xu
2022 arXiv   pre-print
The proposed system is online by design and achieves great robustness and accuracy via: (i) a robustified hybrid SLAM that incorporates learning-based optical flow and/or depth; (ii) self-supervised losses  ...  This dataset is particularly difficult for monocular systems as it contains a large amount of motion blur and rolling-shutter distortion caused by fast camera motion.  ...  We rectify the images with the provided intrinsics to remove image distortion.  ... 
arXiv:2205.01656v1 fatcat:flfov52ebva6dmu3y2s4sdzc24

Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Art [article]

Joel Janai, Fatma Güney, Aseem Behl, Andreas Geiger
2021 arXiv   pre-print
Our survey includes both the historically most relevant literature as well as the current state of the art on several specific topics, including recognition, reconstruction, motion estimation, tracking  ...  As with any rapidly growing field, it becomes increasingly difficult to stay up-to-date or enter the field as a beginner.  ...  They, therefore, propose to explicitly correct the calibration of the camera by exploiting ground truth motion which they use to recover a deformation field by optimizing the reprojection error of point  ... 
arXiv:1704.05519v3 fatcat:xiintiarqjbfldheeg2hsydyra

Area Chairs

Adrian Clark, Alessio Bue, Ist Genova, Andrea Cavallaro, Queen Mary, Andrea Torsello, Andrew Fitzgibbon, Antonio Robles, Kelly Nicta -Canberra, Barbara Caputo, Bjorn Stenger, Edmond Boyer (+28 others)
Foreword It is our great pleasure to welcome you to York for the 27th British Machine Vision Conference (BMVC).  ...  The paper review process was unchanged from previous BMVCs, and we recruited more than 300 reviewers to process the papers.  ...  Acknowledgment: This work was supported by the Austrian Research Promotion Agency project TOFUSION (FIT-IT Bridge program). Acknowledgments R. Mecca is a Marie Curie fellow of the "INdAM", Italy.  ... 

Event-Based Vision Processing in Deep Neural Networks

Bodo Rückauer
One theme in this thesis is the reduction of these obstacles by providing automated tools to convert DNNs to the spike-domain, and to deploy them on neuromorphic hardware.  ...  Inspired by how the DVS removes spatio-temporal redundancy from video, we apply a dynamic masking scheme to the layers of a DNN to reduce the number of operations during inference.  ...  to the original CNNs.  ... 
doi:10.5167/uzh-200987 fatcat:y2clxptcabh2bklipriywhtx4i


1918 Journal of the Society of Chemical Industry  
Tlie Lubes nre iirrnngccl witliin n ciising which is provided with doors on each side to iillow nccess to the tiiks, the lower ends of which iire connected to R trunk proorided with n screw conveyor for  ...  The tiiks lire ltrovirled at tlieir upper ends with spring  ...  lower shutters.  ... 
doi:10.1002/jctb.5000371908 fatcat:5shovdz67zh7rdppg7cnq6gbja

Artificial Intelligence for Efficient Image-based View Synthesis [article]

Thomas Leimkühler, Universität Des Saarlandes, Universität Des Saarlandes
The computer graphics department in the Max-Planck-Institut für Informatik is a wonderfully inspiring environment, which Hans-Peter Seidel created and maintains with generosity and the desire to provide  ...  Finally, I am deeply grateful to my friends and family, who continually help me discover the most important novel views.  ...  control the amount and shape of the radial distortion.  ... 
doi:10.22028/d291-28379 fatcat:kc7nmjh245h2hopgnguqcagpau

Event cameras: from SLAM to high speed video

Henri Rebecq
Acknowledgements Acknowledgements This research was supported by the National Centre of Competence in Research Robotics (NCCR) and the UZH Forschungskredit.  ...  This allows to synthesize motion-corrected event frames, used subsequently to establish feature tracks. Motion Compensation of Events. We synthesize a motion-corrected event image I k (see Figs.  ...  Inspired by [40, 63] , we propose to locally correct the motion of each event according to its individual time stamp.  ... 
doi:10.5167/uzh-187825 fatcat:rqd5ai45srawrh7twodfeagziu

Image Processing for Quanta Image Sensors

Omar A Elgendy
To acquire an image, the sensor oversamples the space and time to generate a sequence of binary bit maps.  ...  QIS is envisioned to be the next generation image sensor after CCD and CMOS since it enables sub-diffraction-limit pixels without the inherited problems of pixel shrinking.  ...  It can also be referred to the shutter duty cycle because the shutter is opened to collect photons during this proportion of time [122] .  ... 
doi:10.25394/pgs.8429048.v1 fatcat:zhod4upwvncdhbef23ve6ysgnu

The Representation of the Violated Body and its Spectatorship: the Negative Approach (1970s to 2016)

Vita Upeniece
Thus, the context of seeming abundance and importance attributed to the photographic image of violence, prompted me to focus on the 'negative' approach and explore artworks where the image of the vulnerable  ...  This study explores the representation of the violated body and its spectatorship, focusing on the 'negative' approach in the period from the 1970's up to 2016.  ...  I am also very thankful to Signe Dempelsone, Nick Bryson, Myra Calvert and Robert Szymanek for proofreading the final draft.  ... 
doi:10.34737/vw732 fatcat:bzvp2gwupje55bhe43btw2zydi