YOLACT: Real-time Instance Segmentation

ICCV 2019 Daniel BolyaChong ZhouFanyi XiaoYong Jae Lee

We present a simple, fully-convolutional model for real-time instance segmentation that achieves 29.8 mAP on MS COCO at 33.5 fps evaluated on a single Titan Xp, which is significantly faster than any previous competitive approach. Moreover, we obtain this result after training on only one GPU... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK LEADERBOARD
Instance Segmentation COCO minival YOLACT-550 (ResNet-50) mask AP 29.9 # 20
Instance Segmentation COCO test-dev YOLACT (ResNet-50-FPN) mask AP 29.8% # 24
Real-time Instance Segmentation MSCOCO YOLACT MAP 29.8 # 1
AP 24.9 # 1
Frame (fps) 45.5 # 2
Real-time Instance Segmentation MSCOCO YOLACT-550 (ResNet-101-FPN) AP 28.2 # 3
Frame (fps) 33.3 # 3

Methods used in the Paper