InstaBoost: Boosting Instance Segmentation via Probability Map Guided Copy-Pasting

TL;DR

InstaBoost introduces a simple, efficient data augmentation technique for instance segmentation that leverages pixel redundancy and location probability maps to improve Mask R-CNN performance without increasing computational cost.

Contribution

The paper proposes a novel augmentation method using existing annotations and location maps to enhance instance segmentation accuracy.

Findings

Improves Mask R-CNN performance by 1.7 mAP on COCO

Increases Pascal VOC mAP by 3.3 with jittering

Boosts R101-Mask R-CNN from 35.7 to 37.9 mAP

Abstract

Instance segmentation requires a large number of training samples to achieve satisfactory performance and benefits from proper data augmentation. To enlarge the training set and increase the diversity, previous methods have investigated using data annotation from other domain (e.g. bbox, point) in a weakly supervised mechanism. In this paper, we present a simple, efficient and effective method to augment the training set using the existing instance mask annotations. Exploiting the pixel redundancy of the background, we are able to improve the performance of Mask R-CNN for 1.7 mAP on COCO dataset and 3.3 mAP on Pascal VOC dataset by simply introducing random jittering to objects. Furthermore, we propose a location probability map based approach to explore the feasible locations that objects can be placed based on local appearance similarity. With the guidance of such map, we boost the performance of R101-Mask R-CNN on instance segmentation from 35.7 mAP to 37.9 mAP without modifying the backbone or network structure. Our method is simple to implement and does not increase the computational complexity. It can be integrated into the training pipeline of any instance segmentation model without affecting the training and inference efficiency. Our code and models have been released at https://github.com/GothicAi/InstaBoost