FastMask: Segment Multi-scale Object Candidates in One Shot

TL;DR

FastMask is a novel deep learning framework that efficiently segments multi-scale objects in one shot, significantly outperforming existing methods in speed and maintaining high accuracy, suitable for real-time applications.

Contribution

We introduce FastMask, a new segment proposal framework utilizing hierarchical features and specialized modules for fast, scale-tolerant object segmentation in a single inference pass.

Findings

Outperforms state-of-the-art in average recall on MS COCO

Achieves near real-time segmentation at ~13 fps

Outperforms existing methods in speed by 2-5 times

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.