Shape Context: A New Descriptor for Shape Matching and Object Recognition

By Serge J. Belongie, Jitendra Malik, Jan Puzicha. 2000

674 citations recorded by the source corpus.

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

Dynamic Graph CNN for Learning on Point Clouds (2018, methodology): “can distinguish between extrinsic and intrinsic descriptors. Extrinsic descriptors usually are derived from the coordinates of the shape in 3D space and includes classical methods like shape context [Belongie et al. 2001], spin images [Johnson and Hebert 1999], integral features [Manay et al. 2006],”
Bird Species Categorization Using Pose Normalized Deep Convolutional Nets (2014, neutral): “The above transformations Name W (y,w) Solve w∗t p # Pts Translation y = yt +T T = µi−µt |S| ≥ 1 2D Similarity y = sRyt + t R =V diag(1,det(VU>))U>, s = tr(M̄>i RM̄t) tr(M̄>t M̄t) , T = µi− sRµt |S| ≥ 2 2D Affine y = Ayht A = MiM h> t (M h t M h> t ) −1 |S| ≥ 3 Table 1: Computation of warping…”
SIFT Flow: Dense Correspondence across Scenes and Its Applications (2011, neutral): “✦”
A review of multimodal image matching: Methods and applications (2021, background): “Typical points descriptors include shape context (SC) [131], spin images [304], integral volume [305], and point feature histograms [306].”
CPMC: Automatic Object Segmentation Using Constrained Parametric Min-Cuts (2012, background): “ground mask, a bag of local shape contexts [74] computed on”
Constrained parametric min-cuts for automatic object segmentation (2010, methodology): “On VOC2009 we have also run experiments where we have complemented the feature set with additional appearance and shape features — a bag of dense SIFT [19] features on the foreground mask, a bag of Local Shape Contexts [4] on its boundary, and a Pyramid HOG [5] with 3 levels — for a total of 1054 fe”
Image Matching from Handcrafted to Deep Features: A Survey (2020, background): “Shape Context (Belongie et al. 2001)] and differences of edge length to determine node and edge affinities.”
Nonparametric scene parsing: Label transfer via dense scene alignment (2009, methodology): “Traditional approaches to object recognition begin by specifying an object model, such as template matching [28, 5], constellations [9, 7], bags of features [24, 14, 10, 25], or shape models [2, 3, 6], etc.”
Shape Matching and Object Recognition Using Shape Contexts (2002, background): “Parts of this work have appeared in [3, 2] We wish”
Back to the Future: Learning Shape Models from 3D CAD Data (2010, methodology): “Following early uses of CAD models for recognition [4], but in contrast to recent work [14, 15], we chose Learning Object Class Models from 3D Meshes | Michael Stark 14 Learning Object Class Models from 3D Meshes | Michael Stark 14 Figure 2: Comparison of shape representations fed into Shape Context”
3D object classification and retrieval with Spherical CNNs (2017, background): “The Intrinsic Shape Context descriptor (ISC) [18] is an extension of the 2D Shape Context [4] descriptor to triangular meshes.”
Learning Spectral Descriptors for Deformable Shape Correspondence (2014, background): “Classical works in this category include the shape context [6] and spin image [7] descriptors, as well as integral volume descriptors [8], [1] and multiscale local features [9], just to mention a few out of many.”
A boosted classifier tree for hand shape detection (2004, methodology): “We have chosen to represent hands images with shape context features proposed initially by Belongie et al. [1].”
Learning class‐specific descriptors for deformable shapes using localized spectral convolutional networks (2015, background): “Another class of approaches try to bring successful models like SIFT [Low04] or shape context, [BMP00] from images to surfaces [SB11,KBLB12].”
Geodesic Convolutional Neural Networks on Riemannian Manifolds (2015, background): “The success of image descriptors such as SIFT [31], HOG [13], MSER [33], and shape context [2] has led to several generalizations thereof to non- Euclidean domains (see e.g. [49] [14] [24], respectively).”
Similarity Scores Based on Background Samples (2009, background): “Some similarity measures proposed in the past have been hand crafted (e.g., [ 3 ,4]).”
A SIFT descriptor with global context (2005, methodology): “More recently, Mikolajczyk, Zisserman , and Schmid [14] integrate edge-based features with local feature-based recognition using a structure similar to shape contexts [2] for general object-class recognition.”