Estimating Spatial Layout of Rooms using Volumetric Reasoning about Objects and Surfaces

By David C. Lee, Abhinav Kumar Gupta, Martial Hebert, Takeo Kanade. Neural Information Processing Systems, 2010

317 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/63d4018a9882eba91da21052164775095d410f23

How researchers see this paper

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

Citations in the corpus

Understanding Bayesian Rooms Using Composite 3D Object Models (2013, methodology): “Indoor scene understanding from monocular images has received much recent interest, and advancements in estimating the 3D geometry of such environments [2, 7, 11, 17] have enabled several interesting applications.”
Single Image 3D without a Single 3D Image (2015, methodology): “At test time, this local mapping is used on a single image to infer the 3D labels; in other works, it is again presumed that there is such regularity that one can impose even more top-down constraints, such as the Manhattan-world assumption [8], an indoor box model [20, 35], or others [27, 6, 5, 17,”
Zero-Shot Task Transfer (2019, methodology): “rst image is the reference frame of the camera, i.e. green arrow. The second image, with red arrow, is taken after a geometric translation w.r.t ﬁrst image. Blue rectangles show our results. Methd VM [15] EM [44] LN [46] TTNet 6 RN [20] TN [42] TTNt 10 TTNt 20 TTNt LS Keypt. 15.48 11.2 7.64 7.51 6.3”
Scene Parsing by Integrating Function, Geometry and Appearance Models (2013, methodology): “(i) Hedau et al. [12, 13], Wang et al. [21], Lee et al. [17, 16] and Satkin et al. [19] adopted different approaches to model the geometric layout of the background and/or foreground blocks with the Structured SVM (or Latent SVM).”
Ieee Transactions on Pattern Analysis and Machine Intelligence 1 3d Reasoning from Blocks to Stability (background): “In addition, researchers have studied indoor environment reasoning on color images, where the 3D geometric inference can be approximated as a Manhattan World [16] [17] [18] [19].”
Efficient Exact Inference for 3D Indoor Scene Understanding (2012, methodology): “While the complexity is determined directly by the number of candidate layouts when working with the univariate parameterization [6, 7], we obtain a complexity dependence on the dimensionality of the domain of the potentials representing the image features, i.”
Understanding Indoor Scenes Using 3D Geometric Phrases (2013, methodology): “Similarly to [12, 16, 27], we represent the indoor space by the 3D layout of 5 orthogonal faces (floor, ceiling, left, center, and right wall), as in Fig.”
Discrete-Continuous Depth Estimation from a Single Image (2014, methodology): “, box models) have proven effective to estimate the layout of a room [9, 17, 27].”
Designing deep networks for surface normal estimation (2014, methodology): “Our fusion network can be viewed as a form of learned reasoning that replaces previous optimization-based attempts to reconcile evidence [12, 24, 29, 10] from conflicting sources.”
Perceptual Organization and Recognition of Indoor Scenes from RGB-D Images (2013, background): “The problem of three dimensional scene understanding from monocular images has received considerable attention in recent years [11, 8, 10, 20, 19, 25], and different aspects have been explored.”
Factoring Shape, Pose, and Layout from the 2D Image of a 3D Scene (2017, methodology): “Similarly, in CAD retrieval scenarios, one assumes the object [3, 4, 13, 19, 20] or scene [17] can be represented in terms of a pre-determined dictionary of shapes; a great deal of earlier work [12, 18, 27] tackled this with a dictionary of box models.”
Efficient Structured Prediction with Latent Variables for General Graphical Models (2012, methodology): “Following existing approaches (Hedau et al., 2009; Lee et al., 2010), we employ F = 55 features based on geometric context (GC) and orientation maps (OM) and refer the interested reader to (Hoiem et al., 2007) and (Lee et al., 2009) for respective details.”
Marr Revisited: 2D-3D Alignment via Surface Normal Prediction (2016, methodology): “Parametric models, such as volumetric models [24], cuboids [48], joint cuboid and room layout [38], and support surfaces (in RGB-D) [13] have been proposed.”
Robust real-time visual odometry for dense RGB-D mapping (2013, methodology): “Although there are several easy-to-use tools for annotating 2D images and videos [18, 26, 22], they would be suboptimal to use for RGB-D sequences because they do not exploit the 3D structure of the problem in label propagation.”