Multi-Class Segmentation with Relative Location Prior

By Stephen Gould, Jim Rodgers, David S. Cohen, Gal Elidan, Daphne Koller. International Journal of Computer Vision, 2008

463 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/599c8d460575ddfea702075b8ccde01b6fe987e8

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

Citations in the corpus

Learning Where to Classify in Multi-view Semantic Segmentation (2014, background): “Additional works included higher-order terms [15, 16] and simplification by superpixels [17, 18].”
Superpixels and Polygons Using Simple Non-iterative Clustering (2017, background): “After their introduction [27], several applications such as object localization [14], multi-class segmentation [15], optical flow [22], body model estimation [24], object tracking [35], and depth estimation [37] took advantage of superpixels .”
Factorizable Net: An Efficient Subgraph-based Framework for Scene Graph Generation (2018, background): “…phrases detection and detection works and these works were mostly supposed to leveraging other tasks, such as object recognition [4,12,13,31,32,44,51,53,55], image classiﬁcation and retrieval [17,39], scene understanding and generation [2,3,18,23,24,61,65], as well as text grounding [27, 43, 49].”
Spatial Pattern Templates for Recognition of Objects with Regular Structure (2013, methodology): “[4] resort to a voting scheme and use CRF with pairwise terms for directly adjacent segments only.”
Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials (2011, background): “Basic CRF models are composed of unary potentials on individual pixels or image patches and pairwise potentials on neighboring pixels or patches [19, 23, 7, 5].”
Harmony potentials for joint classification and segmentation (2010, background): “One of the most successful trends in object class image segmentation poses this labeling problem as one of energy minimization of a conditional random field (CRF) [20, 6, 22].”
SLIC Superpixels ? (2010, methodology): “Our first task is to perform object class recognition for 21 object classes from the STAIR vision library(5) based on the work of Gould et al [22].”
Superpixels: An evaluation of the state-of-the-art (2016, methodology): “Similarly, the normalized cuts algo-rithm was early adopted for oversegmentation [1] and semantic segmentation [8].”
Exploring Visual Relationship for Image Captioning (2018, background): “Some early works [9,10] attempt to learn four spatial relations (i.”
Stacked Hierarchical Labeling (2010, methodology): “[7]∗ 77 64 72 95 81 66 71 93 74 70 70 69 72 68 55 23 83 40 77 60 50 50 14 [31] 75 65 77 93 70 58 64 92 57 70 61 69 67 74 70 47 80 53 73 53 56 47 40 [21] – 67 30 71 69 68 64 84 88 58 77 82 91 90 82 34 93 74 31 56 54 54 49 [27] 75 69 69 96 87 78 80 95 83 67 84 70 79 47 61 30 80 45 78 68 52 67 27 [20] ”
Class segmentation and object localization with superpixel neighborhoods (2009, background): “Class segmentation algorithms which operate at the pixel level are often based on local features like textons [32] and are augmented by a conditional random field or another spatial coherency aid [15, 19, 16, 37, 17, 28, 13] to refine the results.”
Effective Use of Synthetic Data for Urban Scene Semantic Segmentation (2018, background): “Semantic segmentation, that is, understanding an image at pixel-level, has been widely studied by the computer vision community [43, 46, 15, 26, 11, 33, 42, 25, 28, 4, 53, 52].”
Visual Relationship Detection with Language Priors (2016, background): “Some previous work has attempted to learn spatial relationships between objects [19,13] to improve segmentation [19].”
SLIC Superpixels Compared to State-of-the-Art Superpixel Methods (2012, methodology): “26 GHz processor with 2GB RAM, and the class-averaged segmentation accuracy obtained on the MSRC data set using the method described in [11].”
Ieee Transactions on Pattern Analysis and Machine Intelligence 1 Exploring Tiny Images: the Roles of Appearance and Contextual Information for Machine and Human Object Recognition (unknown): “Gould et al. [52] learn the relative locations of objects and use this information to improve upon appearance-based segmentation.”
Pooling Methods in Deep Neural Networks, a Review (2020, background): “Owing to its representational and computational efficiency, super-pixel has become a midlevel image representation and is widely used in computer vision algorithms such as object detection [29, 30], semantic segmentation [31-34], saliency estimation [35-38], optical flow estimation [39, 40], depth e”
Learning to segment moving objects in videos (2014, background): “…window classiﬁers [19] that often cannot afford to enumerate all possible bounding boxes in an image, or Markov Random Field pixel clas-siﬁers that make independence assumptions regarding the organization of pixel labels, and cannot distinguish closeby instances of the same object class [9,16].”