Path Aggregation Network for Instance Segmentation

• Published in: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition pp. 8759 - 8768
• Main Authors: Liu, Shu, Qi, Lu, Qin, Haifang, Shi, Jianping, Jia, Jiaya
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2018
• Subjects: Feature extraction; Image segmentation; Object detection; Proposals; Semantics; Task analysis; Training

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. Specifically, we enhance the entire feature hierarchy with accurate localiz...