FastMask: Segment Multi-Scale Object Candidates in One Shot

Hexiang Hu, Shiyi Lan, Yuning Jiang, Zhimin Cao, Fei Sha; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 991-999

Abstract


Objects appear to scale differently in natural images. This fact requires methods dealing with object-centric tasks (e.g. object proposal) to have robust performance over variances in object scales. In the paper, we present a novel segment proposal framework, namely FastMask, which takes advantage of hierarchical features in deep convolutional neural networks to segment multi-scale objects in one shot. Innovatively, we adapt segment proposal network into three different functional components (body, neck and head). We further propose a weight-shared residual neck module as well as a scale-tolerant attentional head module for efficient one-shot inference. On MS COCO benchmark, the proposed FastMask outperforms all state-of-the-art segment proposal methods in average recall being 2 5 times faster. Moreover, with a slight trade-off in accuracy, FastMask can segment objects in near real time ( 13 fps) with 800*600 resolution images, demonstrating its potential in practical applications. Our implementation is available on https://github.com/voidrank/FastMask.

Related Material


[pdf] [Supp]
[bibtex]
@InProceedings{Hu_2017_CVPR,
author = {Hu, Hexiang and Lan, Shiyi and Jiang, Yuning and Cao, Zhimin and Sha, Fei},
title = {FastMask: Segment Multi-Scale Object Candidates in One Shot},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {July},
year = {2017}
}