Filters








1,032 Hits in 4.4 sec

FlowNet: Learning Optical Flow with Convolutional Networks [article]

Philipp Fischer, Alexey Dosovitskiy, Eddy Ilg, Philip Häusser, Caner Hazırbaş, Vladimir Golkov, Patrick van der Smagt, Daniel Cremers, Thomas Brox
2015 arXiv   pre-print
In this paper we construct appropriate CNNs which are capable of solving the optical flow estimation problem as a supervised learning task.  ...  Optical flow estimation has not been among the tasks where CNNs were successful.  ...  Supplementary Material for 'FlowNet: Learning Optical Flow with Convolutional Networks' Flow field color coding To visualize the flow fields, we use the tool provided with Sintel [7] .  ... 
arXiv:1504.06852v2 fatcat:po46vuso2renfdy7hnivq5g7zy

FlowNet: Learning Optical Flow with Convolutional Networks

Alexey Dosovitskiy, Philipp Fischer, Eddy Ilg, Philip Hausser, Caner Hazirbas, Vladimir Golkov, Patrick van der Smagt, Daniel Cremers, Thomas Brox
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
In this paper we construct appropriate CNNs which are capable of solving the optical flow estimation problem as a supervised learning task.  ...  Optical flow estimation has not been among the tasks where CNNs were successful.  ...  Supplementary Material for 'FlowNet: Learning Optical Flow with Convolutional Networks' Flow field color coding To visualize the flow fields, we use the tool provided with Sintel [7] .  ... 
doi:10.1109/iccv.2015.316 dblp:conf/iccv/DosovitskiyFIHH15 fatcat:mehno2azlbcmxlhvthtnpdqzni

EV-FlowNet: Self-Supervised Optical Flow Estimation for Event-based Cameras

Alex Zhu, Liangzhe Yuan, Kenneth Chaney, Kostas Daniilidis
2018 Robotics: Science and Systems XIV  
To these points, we present EV-FlowNet, a novel self-supervised deep learning pipeline for optical flow estimation for event based cameras.  ...  We show that the resulting network is able to accurately predict optical flow from events only in a variety of different scenes, with performance competitive to image based networks.  ...  These event images combined with the self-supervised loss are sufficient for the network to learn to predict accurate optical flow from events alone.  ... 
doi:10.15607/rss.2018.xiv.062 dblp:conf/rss/ZhuYCD18 fatcat:ondxeg7pxjaffk3mr5e44ljkpa

FlowNet-PET: Unsupervised Learning to Perform Respiratory Motion Correction in PET Imaging [article]

Teaghan O'Briain, Carlos Uribe, Kwang Moo Yi, Jonas Teuwen, Ioannis Sechopoulos, Magdalena Bazalova-Carter
2022 arXiv   pre-print
The network was trained to predict the optical flow between two PET frames from different breathing amplitude ranges.  ...  To correct for respiratory motion in PET imaging, an interpretable and unsupervised deep learning technique, FlowNet-PET, was constructed.  ...  Optical flow estimation can be accomplished through supervised learning with convolutional neural networks (CNNs) (Fischer et al., 2015; Teed and Deng, 2020) .  ... 
arXiv:2205.14147v3 fatcat:o36rnj3dpzbdliviefwh5jk4cq

Spike-FlowNet: Event-based Optical Flow Estimation with Energy-Efficient Hybrid Neural Networks [article]

Chankyu Lee, Adarsh Kumar Kosta, Alex Zihao Zhu, Kenneth Chaney, Kostas Daniilidis, Kaushik Roy
2020 arXiv   pre-print
The network is end-to-end trained with self-supervised learning on Multi-Vehicle Stereo Event Camera (MVSEC) dataset.  ...  To overcome these issues, we present Spike-FlowNet, a deep hybrid neural network architecture integrating SNNs and ANNs for efficiently estimating optical flow from sparse event camera outputs without  ...  Authors of [27] presented a methodology for optical flow estimation using convolutional SNNs based on Spike-Time-Dependent-Plasticity (STDP) learning [11] .  ... 
arXiv:2003.06696v3 fatcat:e5k5lxzh2veqrdllpluetcydyi

3D-FlowNet: Event-based optical flow estimation with 3D representation [article]

Haixin Sun, Minh-Quan Dao, Vincent Fremont
2022 arXiv   pre-print
We then propose 3D-FlowNet, a novel network architecture that can process the 3D input representation and output optical flow estimations according to the new encoding methods.  ...  The results show that our 3D-FlowNet outperforms state-of-the-art approaches with less training epoch (30 compared to 100 of Spike-FlowNet).  ...  [16] proposed a network that can learn optical flow from brightness constancy and motion smoothness. Based on that, Meister et al.  ... 
arXiv:2201.12265v1 fatcat:jwbrvtprd5e4jonzifr5prhp6q

FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks [article]

Eddy Ilg, Nikolaus Mayer, Tonmoy Saikia, Margret Keuper, Alexey Dosovitskiy, Thomas Brox
2016 arXiv   pre-print
The FlowNet demonstrated that optical flow estimation can be cast as a learning problem.  ...  Moreover, we present faster variants that allow optical flow computation at up to 140fps with accuracy matching the original FlowNet.  ...  Following FlowNet, several papers have studied optical flow estimation with CNNs: featuring a 3D convolutional network [31] , an unsupervised learning objective [1, 34] , carefully designed rotationally  ... 
arXiv:1612.01925v1 fatcat:upqrnsois5fkta24rxlzyujvy4

FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

Eddy Ilg, Nikolaus Mayer, Tonmoy Saikia, Margret Keuper, Alexey Dosovitskiy, Thomas Brox
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
The FlowNet demonstrated that optical flow estimation can be cast as a learning problem.  ...  Moreover, we present faster variants that allow optical flow computation at up to 140fps with accuracy matching the original FlowNet.  ...  The idea of using a simple convolutional neural network (CNN) architecture to directly learn the concept of optical flow from data was completely disjoint from all the established approaches.  ... 
doi:10.1109/cvpr.2017.179 dblp:conf/cvpr/IlgMSKDB17 fatcat:ffxpoiw67vaufazauyniwnyzqi

Fusion-FlowNet: Energy-Efficient Optical Flow Estimation using Sensor Fusion and Deep Fused Spiking-Analog Network Architectures [article]

Chankyu Lee, Adarsh Kumar Kosta, Kaushik Roy
2021 arXiv   pre-print
To address such issues associated with the sensors, we present Fusion-FlowNet, a sensor fusion framework for energy-efficient optical flow estimation using both frame- and event-based sensors, leveraging  ...  Our network is end-to-end trained using unsupervised learning to avoid expensive video annotations.  ...  Figure 5 . 5 Predicted optical flow compared with other state-of-the-art methods. EV-FlowNet [32] and Spike-FlowNet [15] use only the event stream as input.  ... 
arXiv:2103.10592v1 fatcat:livjsh4jizft7can4jxb6hwhkm

HPLFlowNet: Hierarchical Permutohedral Lattice FlowNet for Scene Flow Estimation on Large-Scale Point Clouds

Xiuye Gu, Yijie Wang, Chongruo Wu, Yong Jae Lee, Panqu Wang
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
We present a novel deep neural network architecture for end-to-end scene flow estimation that directly operates on large-scale 3D point clouds.  ...  Our model can efficiently process a pair of point cloud frames at once with a maximum of 86K points per frame.  ...  It is the 3D counterpart of optical flow, and is a more fundamental and unambiguous representation -optical flow is simply the projection of scene flow onto the image plane of a camera [42] .  ... 
doi:10.1109/cvpr.2019.00337 dblp:conf/cvpr/GuWWLW19 fatcat:dmurypc5r5fknkoslwlqdskxci

Finding Correspondences for Optical Flow and Disparity Estimations using a Sub-pixel Convolution-based Encoder-Decoder Network [article]

Juan Luis Gonzalez, Muhammad Sarmad, Hyunjoo J.Lee, Munchurl Kim
2018 arXiv   pre-print
In this paper, we propose a novel sub-pixel convolution-based encoder-decoder network for optical flow and disparity estimations, which can extend FlowNetS and DispNet by replacing the deconvolution layers  ...  with sup-pixel convolution blocks.  ...  [4] have used a CNN-based optical flow estimation network, called the FlowNet.  ... 
arXiv:1810.03155v1 fatcat:pr7uc75tgza4bgdmummd5rfvum

Optical Flow Estimation Using a Spatial Pyramid Network

Anurag Ranjan, Michael J. Black
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We learn to compute optical flow by combining a classical spatial-pyramid formulation with deep learning.  ...  Our results are more accurate than FlowNet on most standard benchmarks, suggesting a new direction of combining classical flow methods with deep learning.  ...  Acknowledgement We thank Max Planck ETH Center for Learning Systems for their support. We thank Jonas Wulff for his insightful discussions about optical flow.  ... 
doi:10.1109/cvpr.2017.291 dblp:conf/cvpr/RanjanB17 fatcat:dyjoerjg4bbmdmcva5utwlqob4

Optical Flow Estimation using a Spatial Pyramid Network [article]

Anurag Ranjan, Michael J. Black
2016 arXiv   pre-print
We learn to compute optical flow by combining a classical spatial-pyramid formulation with deep learning.  ...  Our results are more accurate than FlowNet on most standard benchmarks, suggesting a new direction of combining classical flow methods with deep learning.  ...  Acknowledgement We thank Jonas Wulff for his insightful discussions about optical flow.  ... 
arXiv:1611.00850v2 fatcat:mkel6t3zjnd7dcwvzhbdbd2ypm

Optical Flow Super-Resolution Based on Image Guidence Using Convolutional Neural Network [article]

Liping Zhang, Zongqing Lu, Qingmin Liao
2018 arXiv   pre-print
The convolutional neural network model for optical flow estimation usually outputs a low-resolution(LR) optical flow field.  ...  With the motivation of various convolutional neural network(CNN) structures succeeded in single image super-resolution(SISR) task, an end-to-end convolutional neural network is proposed to reconstruct  ...  Compared with SISR, the optical flow SR network is able to obtain more information for learning.  ... 
arXiv:1809.00588v1 fatcat:7dmv3yx5hnfyjppib2wkxidldu

ActionFlowNet: Learning Motion Representation for Action Recognition [article]

Joe Yue-Hei Ng, Jonghyun Choi, Jan Neumann, Larry S. Davis
2018 arXiv   pre-print
We propose a multitask learning model ActionFlowNet to train a single stream network directly from raw pixels to jointly estimate optical flow while recognizing actions with convolutional neural networks  ...  We additionally provide insights to how the quality of the learned optical flow affects the action recognition.  ...  Multi-frame Optical Flow with 3D-ResNet Fischer et al. proposed FlowNet [7] that is based on convolutional neural networks to estimate high quality optical flow.  ... 
arXiv:1612.03052v3 fatcat:priemyeu7jbx5jyp7a7xj2odzu
« Previous Showing results 1 — 15 out of 1,032 results