//

Convolutional Oriented Boundaries

State-of-the-Art Contour Detection and Hierarchical Segmentation


Publications

PAMI

K.K. Maninis, J. Pont-Tuset, P. Arbeláez, and L. Van Gool
Convolutional Oriented Boundaries: From Image Segmentation to High-Level Tasks
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2017.
[PDF] [BibTex]

@article{Man+17,
author = {K.K. Maninis and J. Pont-Tuset and P. Arbel\'{a}ez and L. Van Gool},
title = {Convolutional Oriented Boundaries: From Image Segmentation to High-Level Tasks},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)},
year = {2017}
}
ECCV

K.K. Maninis, J. Pont-Tuset, P. Arbeláez, and L. Van Gool
Convolutional Oriented Boundaries
European Conference on Computer Vision (ECCV) 2016
[PDF] [BibTex]

@inproceedings{Man+16a,
author = {K.K. Maninis and J. Pont-Tuset and P. Arbel\'{a}ez and L. Van Gool},
title = {Convolutional Oriented Boundaries},
booktitle = {European Conference on Computer Vision (ECCV)},
year = {2016}
}
Please cite these papers if you found the resources of this web useful.

Abstract

This paper presents Convolutional Oriented Boundaries (COB), which produces multiscale oriented contours and region hierarchies starting from generic image classification CNNs. COB is computationally efficient, because it requires a single CNN forward pass for contour detection and 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. 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.

Demo of the Detected Orientations, UCM, and Object Proposal

Top row: Contour response at different orientations. Bottom row: Hierarchical segmentation (UCM) and best segmented object proposal.

Benchmark State of the Art

Display the quantitative evaluation of the current State-of-the-Art techniques in PASCAL Context.

Explore State of the Art

Qualitatively compare current State-of-the-Art techniques in PASCAL Context.

Code & Pre-Computed Results

Download the COB code and pre-computed COB segmentations and proposals in a variety of popular datasets: BSDS500, MS-COCO, PASCAL, etc.