Filters








9 Hits in 5.3 sec

CAM-Convs: Camera-Aware Multi-Scale Convolutions for Single-View Depth [article]

Jose M. Facil, Benjamin Ummenhofer, Huizhong Zhou, Luis Montesano, Thomas Brox, Javier Civera
2019 arXiv   pre-print
Single-view depth estimation suffers from the problem that a network trained on images from one camera does not generalize to images taken with a different camera model.  ...  In this work, we propose a new type of convolution that can take the camera parameters into account, thus allowing neural networks to learn calibration-aware patterns.  ...  We also thank Facebook for their P100 server donation and gift funding; and Nvidia for their Titan X and Xp donation.  ... 
arXiv:1904.02028v1 fatcat:rgsnk4ck6fbfhbmjsanpycxceq

CAM-Convs: Camera-Aware Multi-Scale Convolutions for Single-View Depth

Jose M. Facil, Benjamin Ummenhofer, Huizhong Zhou, Luis Montesano, Thomas Brox, Javier Civera
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Single-view depth estimation suffers from the problem that a network trained on images from one camera does not generalize to images taken with a different camera model.  ...  In this work, we propose a new type of convolution that can take the camera parameters into account, thus allowing neural networks to learn calibration-aware patterns.  ...  We also thank Facebook for their P100 server donation and gift funding; and Nvidia for their Titan X and Xp donation.  ... 
doi:10.1109/cvpr.2019.01210 dblp:conf/cvpr/FacilUZMBC19 fatcat:vp7lhtkhwfgkrleozuo5ogqfu4

Self-Supervised Monocular Scene Flow Estimation [article]

Junhwa Hur, Stefan Roth
2020 arXiv   pre-print
By taking an inverse problem view, we design a single convolutional neural network (CNN) that successfully estimates depth and 3D motion simultaneously from a classical optical flow cost volume.  ...  Our model achieves state-of-the-art accuracy among unsupervised/self-supervised learning approaches to monocular scene flow, and yields competitive results for the optical flow and monocular depth estimation  ...  Upon the augmentation, we also explore the recent CAM-Convs [9] , which facilitate depth estimation irrespective of the camera intrinsics.  ... 
arXiv:2004.04143v2 fatcat:yfej2jtjh5gg5nqkm2tjivde7a

UnRectDepthNet: Self-Supervised Monocular Depth Estimation using a Generic Framework for Handling Common Camera Distortion Models [article]

Varun Ravi Kumar, Senthil Yogamani, Markus Bach, Christian Witt, Stefan Milz, Patrick Mader
2020 arXiv   pre-print
In this paper, we propose a generic scale-aware self-supervised pipeline for estimating depth, euclidean distance, and visual odometry from unrectified monocular videos.  ...  In classical computer vision, rectification is an integral part of multi-view depth estimation. It typically includes epipolar rectification and lens distortion correction.  ...  We want to thank Ciarán Eising (Valeo) and Ravi Kiran (Navya) for providing a detailed review.  ... 
arXiv:2007.06676v3 fatcat:xn7anwto2bgchomdi2cjst66t4

Deep 3D Pan via adaptive "t-shaped" convolutions with global and local adaptive dilations [article]

Juan Luis Gonzalez Bello, Munchurl Kim
2019 arXiv   pre-print
In particular, the generation of new images at parallel camera views given a single input image is of great interest, as it enables 3D visualization of the 2D input scenery.  ...  However, solving the single-image-based view synthesis is still an open problem.  ...  It should be noted that the work of Facil et al. (2019) only handled the supervised monocular depth estimation task for multiple datasets with different camera intrinsics utilizing "CAM-Convs", which  ... 
arXiv:1910.01089v3 fatcat:thfflqsambcl7bz5in2gxu6msy

OmniDet: Surround View Cameras based Multi-task Visual Perception Network for Autonomous Driving [article]

Varun Ravi Kumar, Senthil Yogamani, Hazem Rashed, Ganesh Sistu, Christian Witt, Isabelle Leang, Stefan Milz, Patrick Mäder
2021 arXiv   pre-print
Surround View fisheye cameras are commonly deployed in automated driving for 360 near-field sensing around the vehicle.  ...  We obtain the state-of-the-art results on KITTI for depth estimation and pose estimation tasks and competitive performance on the other tasks.  ...  The closest work is CAM-Convs [30] , which uses camera-aware convolutions for pinhole cameras.  ... 
arXiv:2102.07448v2 fatcat:tw53mpu26ndefltbn5zkaspc7u

On the Uncertain Single-View Depths in Endoscopies [article]

Javier Rodríguez-Puigvert, David Recasens, Javier Civera, Rubén Martínez-Cantín
2021 arXiv   pre-print
In this paper, we explore for the first time Bayesian deep networks for single-view depth estimation in colonoscopies.  ...  As the domain specificity of colonoscopies -- a deformable low-texture environment with fluids, poor lighting conditions and abrupt sensor motions -- pose challenges to multi-view approaches, single-view  ...  Civera, “CAM-Convs: camera-aware multi-scale convolutions for analysis, vol. 71, p. 102058, 2021.  ... 
arXiv:2112.08906v1 fatcat:qcyhq4xwdjf7fbtws5gkjr6w7m

Defocus Map Estimation and Deblurring from a Single Dual-Pixel Image [article]

Shumian Xin, Neal Wadhwa, Tianfan Xue, Jonathan T. Barron, Pratul P. Srinivasan, Jiawen Chen, Ioannis Gkioulekas, Rahul Garg
2021 arXiv   pre-print
Our method is inspired from recent works that leverage the dual-pixel sensors available in many consumer cameras to assist with autofocus, and use them for recovery of defocus maps or all-in-focus images  ...  We use data captured with a consumer smartphone camera to demonstrate that, after a one-time calibration step, our approach improves upon prior works for both defocus map estimation and blur removal, despite  ...  We thank David Salesin and Samuel Hasinoff for helpful feedback. S.X. and I.G. were supported by NSF award 1730147 and a Sloan Research Fellowship.  ... 
arXiv:2110.05655v1 fatcat:gh4vmrmhwna2tem2ww5nxdmxni

Intrinsic motivation mecanisms for incremental learning of visual saliency

Celine Craye, Celine
2017 unpublished
, devant, derrière ou sous une caméra, au fond d'un train 1 ou dans un vieux grenier 2 .  ...  Intrinsic motivation mecanisms for incremental learning of visual saliency. Artificial Intelligence [cs.AI]. Université Paris-Saclay, 2017.  ...  The foveal cameras are replaced by a single EXG50 Baumer camera 4 with a narrow field of view of 5 • .  ... 
fatcat:a2jqued3wbbebhitxeis5m554q