Support Surface Prediction in Indoor Scenes

By Ruiqi Guo, Derek Hoiem. 2013

94 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Geology

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 Geology

Citations in the corpus

ScanNet: Richly-Annotated 3D Reconstructions of Indoor Scenes (2017, background): “The semantic seg-mentations are more than an order of magnitude larger than any previous RGB-D dataset.”
Marr Revisited: 2D-3D Alignment via Surface Normal Prediction (2016, methodology): “Specifically, we can train a 2D → 2.5D model using RGB-D data, and whose output can be aligned to an extracted 2.5D representation of the CAD models.”
Scene Understanding by Reasoning Stability and Safety (2015, background): “…occlusion, and support relations in outdoor scenes inspired by physical reasoning from a 2D image, and Silberman et al. (2012) (Jiang et al. 2013; Guo and Hoiem 2013) who infers the support relations between objects from a single depth image using supervised learning with many prior features.”
Single Image 3D without a Single 3D Image (2015, background): “In this paper, we take this argument one step further: we claim that there is enough regularity in indoor scenes to learn a model for 3D scene understanding without ever seeing an explicit 3D label.”
Generative Image Modeling Using Style and Structure Adversarial Networks (2016, methodology): “We use the 3D models annotated in [57] to generate the synthetic scenes.”
Aligning 3D models to RGB-D images of cluttered scenes (2015, methodology): “Moreover, our algorithm does not rely on detailed annotations (which take about 5 minutes for each scene) [9] of the form used in [10].”
Three-Dimensional Object Detection and Layout Prediction Using Clouds of Oriented Gradients (2016, background): “Recent advances in depth sensor technology have greatly reduced the ambiguities present in standard RGB images, enabling breakthroughs in scene layout prediction [22, 13, 41], support surface prediction [34, 8, 10], semantic parsing [11], and object detection [36].”
Structured Prediction of Unobserved Voxels from a Single Depth Image (2016, methodology): “We use our model trained on the synthetic dataset of [19] from the previous section.”
Learning Rich Features from RGB-D Images for Object Detection and Segmentation (2014, methodology): “’s [17] annotations replace all non-furniture objects with cuboids, changing the statistics of the generated images.”
Unfolding an Indoor Origami World (2014, background): “Indeed, on these scenes, it turns out to be difficult even with depth to find a variety of simple primitives such as cuboids [16,17,39], support surfaces [7], or segmentations [9].”
RGBD Datasets: Past, Present and Future (2016, methodology): “A synthetic re-creation of the 3D scenes also exists [41] ’12”
Dark, Beyond Deep: A Paradigm Shift to Cognitive AI with Humanlike Common Sense (2020, background): “In the next few years, other work [222, 223, 224, 225, 226, 109, 32, 227, 228, 34] also integrated the inference of physical relationships for various scene understanding tasks.”
Inferring 3D Object Pose in RGB-D Images (2015, methodology): “Guo and Hoiem annotate the NYUD2 dataset with 3D CAD models for the following 6 categories: chair, bed, sofa, table, desk and book shelf.”
T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects (2017, methodology): “The UBC VRS [32], the RMRC (a subset of NYU Depth v2 [41] with annotations derived from [18]), the B3DO [26], and the SUN RGB-D [43] provide no 3D models and ground truth only in the form of bounding boxes.”
Joint 3D Object and Layout Inference from a Single RGB-D Image (2015, background): “1 and leverage the manually annotated 3D object ground truth of [11].”
Sliding Shapes for 3D Object Detection in Depth Images (2014, methodology): “RGB-D Scene Understanding: Besides the RGB-D segmentation and detection works mentioned above, [83, 19, 84] proposed to estimate the room layout, support surfaces, and scene understanding for the whole room including objects.”
A Robust 3D-2D Interactive Tool for Scene Segmentation and Annotation (2016, background): “Examples of this approach can be found in [17], [18], [23].”