Instance-aware Semantic Segmentation via Multi-task Network Cascades

CVPR 2016 Jifeng DaiKaiming HeJian Sun

Semantic segmentation research has recently witnessed rapid progress, but many leading methods are unable to identify object instances. In this paper, we present Multi-task Network Cascades for instance-aware semantic segmentation... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
Instance Segmentation COCO test-dev MNC AP50 44.3% # 20
AP75 24.8% # 16
APS 4.7% # 20
APM 25.9% # 18
APL 43.6% # 15
Multi-Human Parsing PASCAL-Person-Part MNC AP 0.5 38.80% # 3

Methods used in the Paper