Convolutional Oriented Boundaries
This work addresses the need for efficient and accurate contour detection and region hierarchies in computer vision, with broad applications in image analysis and object recognition, representing a significant leap over prior methods.
The paper tackles the problem of contour detection and hierarchical segmentation in images by introducing Convolutional Oriented Boundaries (COB), which uses a single CNN forward pass and a sparse boundary representation to achieve state-of-the-art performance across multiple datasets like BSDS, PASCAL Context, PASCAL Segmentation, and MS-COCO.
We present Convolutional Oriented Boundaries (COB), which produces multiscale oriented contours and region hierarchies starting from generic image classification Convolutional Neural Networks (CNNs). COB is computationally efficient, because it requires a single CNN forward pass for contour detection and it uses a novel sparse boundary representation for hierarchical segmentation; it gives a significant leap in performance over the state-of-the-art, and it generalizes very well to unseen categories and datasets. Particularly, we show that learning to estimate not only contour strength but also orientation provides more accurate results. We perform extensive experiments on BSDS, PASCAL Context, PASCAL Segmentation, and MS-COCO, showing that COB provides state-of-the-art contours, region hierarchies, and object proposals in all datasets.