ML p(r)ior | A-expansion for multiple "hedgehog" shapes

A-expansion for multiple "hedgehog" shapes

2016-02-02
1602.01006 | cs.CV
Overlapping colors and cluttered or weak edges are common segmentation problems requiring additional regularization. For example, star-convexity is popular for interactive single object segmentation due to simplicity and amenability to exact graph cut optimization. This paper proposes an approach to multiobject segmentation where objects could be restricted to separate "hedgehog" shapes. We show that a-expansion moves are submodular for our multi-shape constraints. Each "hedgehog" shape has its surface normals constrained by some vector field, e.g. gradients of a distance transform for user scribbles. Tight constraint give an extreme case of a shape prior enforcing skeleton consistency with the scribbles. Wider cones of allowed normals gives more relaxed hedgehog shapes. A single click and +/-90 degrees normal orientation constraints reduce our hedgehog prior to star-convexity. If all hedgehogs come from single clicks then our approach defines multi-star prior. Our general method has significantly more applications than standard one-star segmentation. For example, in medical data we can separate multiple non-star organs with similar appearances and weak or noisy edges.
PDF

Highlights - Most important sentences from the article

Login to like/save this paper, take notes and configure your recommendations

Related Articles

2016-03-13

Interactive object selection is a very important research problem and has many applications. Previou… show more
PDF

Highlights - Most important sentences from the article

2017-04-28

We propose to constrain segmentation functionals with a dimensionless, unbiased and position-indepen… show more
PDF

Highlights - Most important sentences from the article

2013-09-16

Superpixel algorithms aim to over-segment the image by grouping pixels that belong to the same objec… show more
PDF

Highlights - Most important sentences from the article

2019-03-07

We consider the problem of segmenting a biomedical image into anatomical regions of interest. We spe… show more
PDF

Highlights - Most important sentences from the article

2018-09-07

The simplicity of gradient descent (GD) made it the default method for training ever-deeper and comp… show more
PDF

Highlights - Most important sentences from the article

2019-01-11

In this paper, a multi-scale framework with local region based active contour and boundary shape sim… show more
PDF

Highlights - Most important sentences from the article

2019-03-27

We treat shape co-segmentation as a representation learning problem and introduce BAE-NET, a branche… show more
PDF

Highlights - Most important sentences from the article

2019-02-06

Deformable shape modeling approaches that describe objects in terms of their medial axis geometry (e… show more
PDF

Highlights - Most important sentences from the article

2018-07-31
1807.11847 | cs.GR

We present a simple and efficient method based on deep learning to automatically decompose sketched … show more
PDF

Highlights - Most important sentences from the article

2016-12-08

This paper introduces a deep architecture for segmenting 3D objects into their labeled semantic part… show more
PDF

Highlights - Most important sentences from the article

2018-11-01

Recent thrust in imaging capabilities in medical as well as emerging areas of manufacturing systems … show more
PDF

Highlights - Most important sentences from the article

2018-09-13
1809.04995 | cs.CV

Fully connected pairwise Conditional Random Fields (Full-CRF) with Gaussian edge weights can achieve… show more
PDF

Highlights - Most important sentences from the article

2018-08-27

Recent machine learning strategies for segmentation tasks have shown great ability when trained on l… show more
PDF

Highlights - Most important sentences from the article

2018-07-17

High-level shape understanding and technique evaluation on large repositories of 3D shapes often ben… show more
PDF

Highlights - Most important sentences from the article

2018-07-05

Many computer vision methods use consensus maximization to relate measurements containing outliers w… show more
PDF

Highlights - Most important sentences from the article

2012-08-02

We describe a new optimization scheme for finding high-quality correlation clusterings in planar gra… show more
PDF