Path Aggregation Network for Instance Segmentation

CVPR 2018 โ€ข Shu Liu โ€ข Lu Qi โ€ข Haifang Qin โ€ข Jianping Shi โ€ข Jiaya Jia

The way that information propagates in neural networks is of great importance. In this paper, we propose Path Aggregation Network (PANet) aiming at boosting information flow in proposal-based instance segmentation framework... (read more)

PDF Abstract

Evaluation results from the paper

Task Dataset Model Metric name Metric value Global rank Compare
Instance Segmentation COCO PANet Average Precision 42.0% # 1
Object Detection COCO PANet + ResNeXt-101 Bounding Box AP 47.4 # 2