Perceptual Organization and Recognition of Indoor Scenes from RGB-D Images

By Saurabh Gupta, Pablo Arbeláez, Jitendra Malik. 2013

651 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, 3D Vision & Pose

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

In 3D Vision & Pose

Citations in the corpus

Fully convolutional networks for semantic segmentation (2014, neutral): “[11].”
EDTER: Edge Detection with Transformer (2022, neutral): “[49], gPb+NG [19], SE [15], SE+NG+ [20], OEF [21], SemiContour [69], HED [65], RCF [36], AMH-Net [66], LPCB [12], BDCN [22], and PiDiNet [53].”
Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understanding (2020, methodology): “Using our tool, we annotated our entire dataset of 461 scenes with instance-level NYU40 labels [26, 64].”
Holistically-Nested Edge Detection (2015, methodology): “Here we follow the [6] and perform the experiment on the data set processed by [10]. the NYUD dataset is split into 381 training, 414 validation, and 654 testing images.”
Fast Edge Detection Using Structured Forests (2014, neutral): “F”
Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-scale Convolutional Architecture (2014, methodology): “Barron and Malik [3, 2] infer normals from RGB-D inputs using a set of handcrafted priors, along with illumination and reﬂectance.”
A Multi-World Approach to Question Answering about Real-World Scenes based on Uncertain Input (2014, methodology): “Based on automatic segmentations (AutoSeg, 37 classes, single) (3rd row in Table 3) tests the architecture based on uncertain facts obtained from automatic semantic segmentation [15] where the”
Cross-Stitch Networks for Multi-task Learning (2016, methodology): “For semantic segmentation, we follow the setup from [24] and evaluate on the 40 classes using the standard metrics from their work Setup for Surface Normal Prediction: Following [52], we cast the problem of surface normal prediction as classification into one of 20 categories.”
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation (2015, methodology): “We use the segmentation labels provided in [15] in which labels are processed into 40 classes.”
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation (2015, neutral): “[35] use boundary detection and hierarchi-”
Learning Rich Features from RGB-D Images for Object Detection and Segmentation (2014, methodology): “Semantic Segmentation: Finally, we improve semantic segmentation performance (the task of labeling all pixels with a category, but not differentiating between instances) by using object detections to compute additional features for superpixels in the semantic segmentation system we proposed in [18].”
Designing deep networks for surface normal estimation (2014, methodology): “The edge labels are estimated using the ground-truth depth data in a procedure like [13].”
Semantic Scene Completion from a Single Depth Image (2016, background): “ited to address only part of this problem as shown in Figure 2: RGB-D segmentation approaches consider only visible surfaces without the full 3D shape [6, 26], while shape completion approaches consider only geometry without semantics [3] or a single object out of context [32, 34].”
RefineNet: Multi-path Refinement Networks for High-Resolution Semantic Segmentation (2016, methodology): “We use the segmentation labels provided in [19], in which all labels are mapped to 40 classes.”
Cognitive Mapping and Planning for Visual Navigation (2017, background): “Semantics have also been associated with 3D environments more generally (Koppula et al. 2011; Gupta et al. 2013).”
Matterport3D: Learning from RGB-D Data in Indoor Environments (2017, methodology): “These previous RGB-D datasets have been used to train models for several standard scene understanding tasks, including semantic segmentation [27, 33, 13, 40, 7], 3D object detection [31, 24, 14, 36, 37], normal estimation [41, 23, 9, 3], camera relocalization [30, 39], and others [47, 11, 12].”
Towards a Visual Turing Challenge (2014, methodology): “The current state-of-the-art semantic segmentation methods on the NYU-Depth V2 dataset [50] can discriminate only between up to 37 object categories [2, 60, 61], much fewer to what is needed.”