Abstract
The goal of this paper is to provide an accurate pixel-level segmentation of a deformable foreground ob ject in an image. We com- bine state-of-the-art local image segmentation techniques with a global ob ject-specific contour model to form a coherent energy function over the outline of the ob ject and the pixels inside it. The energy function includes terms from a variant of the TextonBoost method, which labels each pixel as either foreground or background. It also includes terms over landmark points from a LOOPS model [1], which combines global ob ject shape with landmark-specific detectors. We allow the pixel-level segmentation and ob ject outline to inform each other through energy potentials so that they form a coherent ob ject segmentation with glob- ally consistent shape and appearance. We introduce an inference method to optimize this energy that proposes moves within the complex energy space based on multiple initial oversegmentations of the entire image. We show that this method achieves state-of-the-art results in precisely segmenting articulated ob jects in cluttered natural scenes.