Efficient Dense Scene Flow from Sparse or Dense Stereo Data

By Andreas Wedel, Clemens Rabe, Tobi Vaudrey, Thomas Brox, Uwe Franke, Daniel Cremers. 2008

214 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimation (2015, background): “[26] followed the variational framework but decoupled the disparity estimation for larger efficiency and accuracy.”
Kinecting the dots: Particle based scene flow from depth sensors (2011, background): “Some authors [9, 21] attempt to use state of the art depth reconstruction algorithms, to provide or initialise the structure underlying their motion field.”
Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation (2014, background): “Several papers have tackled the problem of estimating the 3D flow-field [28,17,14,2].”
Object scene flow for autonomous vehicles (2015, methodology): “Furthermore, due to the absence of realistic benchmarks with scene flow ground truth, quantitative evaluations are restricted to synthetic images [18, 39] or static scenes such as the ones provided by the Middlebury stereo benchmark [27] or the KITTI stereo and optical flow evaluation [12].”
Stereoscopic Scene Flow Computation for 3D Motion Understanding (2011, methodology): “Currently, variational techniques yield the most accurate optical flow in the literature (Wedel et al. 2008a; Werlberger et al. 2010; Zimmer et al. 2009).”
FlowNet3D: Learning Scene Flow in 3D Point Clouds (2018, background): “Since this seminal work, many others have tried to jointly estimate structure and motion from stereoscopic images [12, 18, 34, 26, 5, 33, 28, 29, 1, 30, 16], mostly in a variational setting with regularizations for smoothness of motion and structure [12, 1, 26], or with assumption of the rigidity of”
3D scene flow estimation with a rigid motion prior (2011, methodology): “To date, most scene flow approaches have relied on relatively simple priors favoring smooth surfaces and motion fields [3, 12, 22].”
Piecewise Rigid Scene Flow (2013, methodology): “Semi-dense ground truth data was acquired by a laser [18] [10] [24] Ours”
An Improved Algorithm for TV-L 1 Optical Flow (2009, background): “One direct application is the computation of scene flow, incorporating three data terms [28].”
SphereFlow: 6 DoF Scene Flow from RGB-D Pairs (2014, methodology): “[29] explicitly decouple the position and velocity estimation steps and estimate dense velocities using a variational approach while reporting frame rates of 5 fps on standard consumer hardware.”
3D Scene Flow Estimation with a Piecewise Rigid Scene Model (2015, methodology): “At each pixel the measurement vector for the ﬁlter is composed of scene ﬂow vectors from the current and the previous frame, which are estimated with the method of Wedel et al. (2008).”
Dense Semi-rigid Scene Flow Estimation from RGBD Images (2014, methodology): “If a stereo or multi-view camera system is available, the scene flow can be computed by enforcing consistency with the observed optical flows [18], by a decoupled [21] or joint [1,8] estimation of structure and motion,”
Multi-view Scene Flow Estimation: A View Centered Variational Approach (2010, background): “, [7, 20, 9], where additional cameras require additional sets of unknowns (e.”
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art (2017, background): “However, unlike Wegner et al. (2013), they allow for arbitrarily shaped segments which adapt to more complex road shapes by searching for putative roads with minimum cost paths based on local features.”