Discriminative learning of Markov random fields for segmentation of 3D scan data

By Dragomir Anguelov, B. Taskar, Vassil Chatalbashev, Daphne Koller, Dinkar Gupta, Geremy Heitz, Andrew Y. Ng. Computer Vision and Pattern Recognition, 2005

433 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computational Mechanics, Geology

View Paper: https://www.semanticscholar.org/paper/55a5e1a4e0068a4f2a8a8bdfbd777c249110ccfe

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 Computational Mechanics

In Geology

Citations in the corpus

Deep Projective 3D Semantic Segmentation (2017, methodology): “Popular alternatives include features based on the 3D structure tensor [7,26,11,1], histogram-based descriptors [7,16,11] such as Spin Images [10] and SHOT [21], and simple color features [26,16,11].”
State of research in automatic as-built modelling (2015, methodology): “Based on this idea and inspired from [9], the authors of [123] train a Conditional Random Field to model the relationships between planar patches extracted from point clouds of buildings.”
Extracting Places and Activities from GPS Traces Using Hierarchical Conditional Random Fields (2007, background): “The reader may notice that discriminative learning can also be performed for generative models such as hidden Markov models (Collins 2002), Kalman filters (Abbeel et al. 2005), or Markov random fields ( Anguelov et al. 2005 ).”
Structured Prediction, Dual Extragradient and Bregman Projections (2006, background): “For example, for the Markov random fields that arise in object segmentation problems in vision (Kumar and Hebert, 2004; Anguelov et al., 2005) the task of finding the most likely assignment reduces to a min-cut problem.”
Online) Subgradient Methods for Structured Prediction (2007, methodology): “(Anguelov et al., 2005) built the maximum margin structured classification problem as a quadratic program (QP) and solved it using CPLEX, a well known commercial solver.”
Sliding Shapes for 3D Object Detection in Depth Images (2014, methodology): “Semantic segmentation: A popular way of formulating object recognition in 3D is to predict the semantic label for each region of a depth map or 3D mesh [37, 38, 11, 10, 39–44].”
Decision tree fields (2011, background): “Most algorithms for learning the parameters of a random field try to learn a class-to-class energy table that does not depend on the image content [1, 2, 20, 32, 33].”
MaskFusion: Real-Time Recognition, Tracking and Reconstruction of Multiple Moving Objects (2018, background): “Semantic scene segmentation: Another line of work [2, 26, 52] aims at segmenting 3D scenes semantically, using Markov Random Fields (MRFs).”
Minimizing Nonsubmodular Functions with Graph Cuts-A Review (2007, methodology): “”) To ensure that the function is submodular, two techniques were employed: 1) enforce submodularity constraint while learning parameters of the energy [3], [14], [27].”
Detection-based object labeling in 3D scenes (2012, background): “Features are either extracted from a fixed neighborhood around 3D points [2], [30] or from small patches generated via an oversegmentation of the scene [25], [21].”
Performance of histogram descriptors for the classification of 3D laser range data in urban environments (2012, methodology): “Both local classifiers are compared with a collective classification approach using Conditional Random Fields (CRF)— the Functional Max-Margin Markov Networks (FM3N) [4].”
Discriminative Models for Multi-Class Object Layout (2009, background): “There has been a wide variety of work in the last few years on contextual modeling in image parsing (Torralba et al. 2004; Sudderth et al. 2005; Hoiem et al. 2008; Galleguillos et al. 2008; Shotton et al. 2006; He et al. 2004; Anguelov et al. 2005).”
Efficient Human Pose Estimation from Single Depth Images (2013, methodology): “Anguelov et al. [28] segment puppets in 3D range scan data into head, limbs, torso, and background using spin images and an MRF. Grest et al. [1] use Iterated Closest Point (ICP) to track a skeleton of a known size and starting position from depth images.”
Shape-based recognition of 3D point clouds in urban environments (2009, background): “Several papers have adapted the machinery of Markov Random Fields to the problem of labeling 3D points [3, 27, 24].”
Kernel Methods in Computer Vision (2009, methodology): “[Anguelov et al., 2005] used structured prediction to segmentation 3D point clouds.”