Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation

By Koichiro Yamaguchi, David A. McAllester, Raquel Urtasun. European Conference on Computer Vision, 2014

359 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/efbafc35347ca68b521848251d5a53abe932800a

How researchers see this paper

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

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

End-to-End Learning of Geometry and Context for Deep Stereo Regression (2017, neutral): “SPS-st [44] 4.”
SegStereo: Exploiting Semantic Information for Disparity Estimation (2018, background): “They either formulate the task as supervised learning [26,22] based on current labeled dataset [14,41],”
Object scene flow for autonomous vehicles (2015, methodology): “Inspired by recent trends in optical flow [21, 30, 40, 41] and stereo [2,3,42], Vogel et al.”
Superpixels: An evaluation of the state-of-the-art (2016, background): “…they have been applied to many important problems in computer vision: tracking [10, 11], stereo and occlusion [12, 13], 3D-reconstruction [14], saliency [15], object detection [16, 17], depth recovery [18] and depth estimation [19], indoor scene understanding [20], scene ﬂow [21], and as basis…”
Efficient Deep Learning for Stereo Matching (2016, methodology): “The KITTI 2012 dataset contains 194 training and 195 test images.”
Fast Optical Flow Using Dense Inverse Search (2016, methodology): “We evaluate against four methods: Semi-Global Matching (SGM) [8], Block Matching (BM) [9], Efficient Large-scale Stereo Matching (ELAS) [10] and Slanted-Plane Stereo (SPS) [11].”
DeepPruner: Learning Efficient Stereo Matching via Differentiable PatchMatch (2019, background): “cost aggregation, semi-global matching [14], and Markov random fields [30, 34], they are able to achieve good performance on relatively simple scenarios.”
A Multi-view Stereo Benchmark with High-Resolution Images and Multi-camera Videos (2017, neutral): “5 bad 1 bad 2 bad 4 avgerr rms A50 A90 A95 A99 SPS-Stereo [44] 57.”
Occlusion Aware Unsupervised Learning of Optical Flow (2017, background): “The last group builds more sophisticated frameworks such as modeling depth or a layered representation of objects to reason about occlusion [50, 48, 58, 43].”
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art (2017, methodology): “Superpixels are used as building blocks for various tasks such as stereo and flow estimation (Yamaguchi et al. (2012, 2013, 2014); Güney & Geiger (2015); Bai et al. (2016)), scene flow (Menze & Geiger (2015); Menze et al. (2015b); Lv et al. (2016)), semantic segmentation (Xiao & Quan (2009);…”
Computing the stereo matching cost with a convolutional neural network (2014, methodology): “A current leading approach includes the slanted plane method (Yamaguchi et al., 2014), where most of the work is done by the stereo method.”
Deep Discrete Flow (2016, methodology): “These approaches can be further categorized into epipolar constrained methods [34, 35], methods which estimate the most likely flow field based on a small set of dense flow field proposals [13,36–38] and methods that estimate flow directly at the pixel level [11, 12,14,39,40].”
Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving (2019, background): “According to (Wang et al., 2019a), depth estimation from stereo pairs of images (Mayer et al., 2016; Yamaguchi et al., 2014; Chang & Chen, 2018) are more accurate than that from monocular (i.e., single) images (Fu et al., 2018; Godard et al., 2017) for 3D object detection.”
Burst photography for high dynamic range and low-light imaging on mobile cameras (2016, methodology): “…and brightness constancy, to more recent techniques that use patches or feature descriptors to construct and “densify” a sparse correspondence [Liu et al. 2011; Brox and Malik 2011], or that use image oversegmentations and directly reason about geometry and occlusion [Yamaguchi et al. 2014].”
Pseudo-LiDAR From Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving (2018, methodology): “We provide results using SPS-STEREO [32] and further analysis on depth estimation in the Supplementary Material.”
Towards Robust CNN-based Object Detection through Augmentation with Synthetic Rain Variations (2019, neutral): “[29] is used.”