SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation

TL;DR

SipMask is a fast, single-stage instance segmentation method that preserves spatial information for improved accuracy and real-time performance, outperforming existing methods on COCO and YouTube-VIS datasets.

Contribution

The paper introduces a novel lightweight spatial preservation module and a mask alignment loss, enhancing accuracy while maintaining high speed in single-stage instance segmentation.

Findings

Outperforms existing single-stage methods on COCO test-dev

Achieves 1.0% higher mask AP than TensorMask with four times faster speed

Provides promising results for real-time video instance segmentation on YouTube-VIS

Abstract

Single-stage instance segmentation approaches have recently gained popularity due to their speed and simplicity, but are still lagging behind in accuracy, compared to two-stage methods. We propose a fast single-stage instance segmentation method, called SipMask, that preserves instance-specific spatial information by separating mask prediction of an instance to different sub-regions of a detected bounding-box. Our main contribution is a novel light-weight spatial preservation (SP) module that generates a separate set of spatial coefficients for each sub-region within a bounding-box, leading to improved mask predictions. It also enables accurate delineation of spatially adjacent instances. Further, we introduce a mask alignment weighting loss and a feature alignment scheme to better correlate mask prediction with object detection. On COCO test-dev, our SipMask outperforms the existing single-stage methods. Compared to the state-of-the-art single-stage TensorMask, SipMask obtains an absolute gain of 1.0% (mask AP), while providing a four-fold speedup. In terms of real-time capabilities, SipMask outperforms YOLACT with an absolute gain of 3.0% (mask AP) under similar settings, while operating at comparable speed on a Titan Xp. We also evaluate our SipMask for real-time video instance segmentation, achieving promising results on YouTube-VIS dataset. The source code is available at https://github.com/JialeCao001/SipMask.