Shape matching and object recognition using shape contexts

By Suhas G. Salve, K. Jondhale. 2010

331 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

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

3D shape segmentation via shape fully convolutional networks (2017, methodology): “In this paper, we combine three features, 22 average geodesic distance (AGD) [41], shape context (SC) [42] 23 and spin image (SI) [43] to produce the shape segmentation.”
Depth-images-based pose estimation using regression forests and graphical models (2015, methodology): “Firstly, we propose a new depth70 shape-based feature, termed 3D Local Shape Context feature (3DLSC), by 71 extending the 2D Shape Context (2DSC) [35] to 3D space, to characterise the 72 location cues between human silhouette and joints.”
Image Matching from Handcrafted to Deep Features: A Survey (2020, background): “Corner features can be defined immediately as the midpoint/endpoint or sparse sampling from an edge or contour (Belongie et al. 2002).”
WarpNet: Weakly Supervised Matching for Single-View Reconstruction (2016, result): “Figure 7 shows a few qualitative matching results comparing the baseline CNN and WarpNet.”
Towards Reversal-Invariant Image Representation (2017, background): “Other examples include Shape Context (Belongie et al. 2002), SURF (Bay et al. 2008),BRIEF (Calonder et al. 2010), BRISK (Leutenegger et al. 2011), ORB (Rublee et al. 2011), FREAK (Alahi et al. 2012), etc. Radial transform (Takacs et al. 2013) and polar analysis (Liu et al. 2014) play important…”
Learning 3D Part Detection from Sparsely Labeled Data (2014, methodology): “We first rewrite our scoring function S(X,Y ) with a joint feature map: s1(x, y) = w T y f(x) = w T s1fs1(x, y) (7) s2(xi, yi, xj , yj) = w T yi,yjα(xi, xj) (8) = wTs2αs2(xi, yi, xj , yj) (9) Letting df be the dimensionality of f(x) (df = 2), then ws1 ∈ R(df∗|Y”
A data driven method for feature transformation (2012, methodology): “When the contours of the object can be successfully extracted, Shape Contexts of Belongie et al. [5], which is a histogram of edge points, with log polar bins around the center of the object, have shown good performance in matching and recognition.”