Displets: Resolving stereo ambiguities using object knowledge

By Fatma Güney, Andreas Geiger. 2015

240 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

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 Media Technology

Citations in the corpus

The Cityscapes Dataset for Semantic Urban Scene Understanding (2016, background): “As no official pixel-wise annotations exist for KITTI, several independent groups have annotated approximately 700 frames [22, 29, 32, 33, 58, 64, 77, 80].”
Stereo Matching by Training a Convolutional Neural Network to Compare Image Patches (2015, methodology): “The method in second place (G¨uney and Geiger, 2015) uses the matching cost computed by our previous work ( ˇZbontar and LeCun, 2015).”
Pyramid Stereo Matching Network (2018, methodology): “In the Displets [8] method, 3D models of vehicles are used to resolve matching ambiguities in reflective and textureless regions.”
End-to-End Learning of Geometry and Context for Deep Stereo Regression (2017, unknown): “9 px 68 Displets v2 [18] 3.”
Learning for Disparity Estimation Through Feature Constancy (2017, unknown): “We further compared our method to several existing methods on the KITTI 2012 and KITTI 2015 benchmarks, including GC-NET [12], L-ResMatch [28], SGM-Net [27], SsSMnet [35], PBCP [26], Displets [5], and MC-CNN [32].”
Cascade Residual Learning: A Two-Stage Convolutional Neural Network for Stereo Matching (2017, background): “Guney and Geiger raised Dis-plets in [8], which utilizes object recognition and semantic segmentation for ﬁnding stereo correspondence.”
Semantic Multi-view Stereo: Jointly Estimating Objects and Voxels (2017, background): “utilize a set of car shapes to improve stereo estimation in urban environments [15].”
Towards Probabilistic Volumetric Reconstruction Using Ray Potentials (2015, background): “Priors expressing spatial smoothness, temporal consistency [12, 34] as well as semantic information [13, 14] can be principally integrated into our probabilistic formulation.”
Learning 3D Shape Completion from Laser Scan Data with Weak Supervision (2018, methodology): “ShapeNet: On ShapeNet, we took 3253 car models, and simplified them using the approach outlined in [22] to obtain watertight meshes.”
DSNet: Joint Learning for Scene Segmentation and Disparity Estimation (2019, unknown): “Displets v2 [15] 3.”
SegStereo: Exploiting Semantic Information for Disparity Estimation (2018, background): “Guney and Geiger [18] leveraged object knowledge in MRF formulation to resolve stereo ambiguity.”
Learning Depth with Convolutional Spatial Propagation Network (2018, methodology): “To validate the algorithm, in addition to comparing over the Scene Flow test set, we also submitted our results to KITTI 2012 and 2015 test evaluation server to compare against other SOTA methods proposed in recent years, including PSM-Net [26], iResNet-i2 [102], GC-Net [29], EdgeStereo [101], SGMNe”
Optical Flow with Semantic Segmentation and Localized Layers (2016, background): “ic segmentation in other low-level vision problems. Object class inﬂuences the way things move, but also inﬂuences their shape. Recent work uses semantic segmentation to resolve ambiguities in stereo [15], to guide 3D reconstruction [16,29], and to constrain the motion of the 3D scene by enforcing c”
Multi-Label Semantic 3D Reconstruction Using Voxel Blocks (2016, unknown): “Another option is to combine geometry and semantics by jointly estimating a depth map and semantic segmentation [17, 8].”
Deep Discrete Flow (2016, background): “While these approaches have shown promising results, their performance still falls considerably behind the state-of-the-art in stereo matching [15, 16].”
Semantic Stereo for Incidental Satellite Images (2018, background): “The two subtasks may be considered separately or sequentially and then fused [26, 27], semantic segmentation may be utilized as a prior for stereo [25, 28], stereo may be utilized as a prior for semantic segmentation [33], or both may be solved together [3, 29].”
Efficient Deep Learning for Stereo Matching (2016, background): “Guney and Geiger [12] investigated the utility of high-level vision tasks such as object recognition and semantic segmentation for stereo matching.”
Joint 3d Estimation of Vehicles and Scene Flow (2015, background): “(G¨uney and Geiger, 2015) leverage semantic information to sample CAD shapes with an application to binocular stereo matching.”
Robust stereo matching with surface normal prediction (2017, background): “[14] proposed to make use of semantic segmentation and object knowledge, especially regular structures of cars, to constrain the disparity and attain good performance for textureless, saturated or semitransparent surfaces.”
Patches, Planes and Probabilities: A Non-Local Prior for Volumetric 3D Reconstruction (2016, methodology): “, by constraining the set of plausible geometries [2, 11, 19] or by modeling class specific normal distributions [20,21].”