Laser & Optoelectronics Progress, Volume. 56, Issue 4, 041502(2019)
Mask R-CNN Object Detection Method Based on Improved Feature Pyramid
Figures & Tables(7)
Table 1. Comparison of mAP results in instance segmentation%
View tableTable 1. Comparison of mAP results in instance segmentation%
Method Backbone P APP AP50P AP75P APsP APmP APlMNC[20] ResNet-101-C4 24.6 44.3 24.8 4.7 25.9 43.6 FCIS[21]+OHEM ResNet-101-C5-dilated 29.2 49.5 - 7.1 31.3 50.0 Mask R-CNN ResNet-101-FPN 35.7 58.0 37.8 15.5 38.1 52.4 Proposed ResNet-101-improved FPN 37.2 58.2 39.4 18.4 45.8 52.7
Table 2. Comparison of mAP results in bounding box%
View tableTable 2. Comparison of mAP results in bounding box%
Method Backbone P APbbP AP50bbP AP75bbP APsbbP APmbbP APlbbFaster R-CNN ResNet-101-C4 34.9 55.7 37.4 15.6 38.7 50.9 Faster R-CNN with FPN ResNet-101-FPN 36.2 59.1 39.0 18.2 39.0 48.2 Mask R-CNN ResNet-101-FPN 38.2 60.3 41.7 20.1 41.1 50.2 Proposed ResNet-101-improved FPN 42.3 61.4 45.6 24.2 49.6 51.3
Tools
Get Citation
Copy Citation Text
Zhijun Ren, Suzhen Lin, Dawei Li, Lifang Wang, Jianhong Zuo. Mask R-CNN Object Detection Method Based on Improved Feature Pyramid[J]. Laser & Optoelectronics Progress, 2019, 56(4): 041502
Paper Information
Category: Machine Vision
Received: Aug. 27, 2018
Accepted: Sep. 4, 2018
Published Online: Jul. 31, 2019
The Author Email: Suzhen Lin (lsz@nuc.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20