Method Details


Details for method 'Panoptic-DeepLab [Mapillary Vistas]'

 

Method overview

name Panoptic-DeepLab [Mapillary Vistas]
challenge instance-level semantic labeling
details We employ a stronger backbone, WR-41, in Panoptic-DeepLab. For Panoptic-DeepLab, please refer to https://arxiv.org/abs/1911.10194. For wide-ResNet-41 (WR-41) backbone, please refer to https://arxiv.org/abs/2005.10266.
publication Panoptic-DeepLab: A Simple, Strong, and Fast Baseline for Bottom-Up Panoptic Segmentation
Bowen Cheng, Maxwell D. Collins, Yukun Zhu, Ting Liu, Thomas S. Huang, Hartwig Adam, Liang-Chieh Chen
https://arxiv.org/abs/1911.10194
project page / code
used Cityscapes data fine annotations
used external data ImageNet, Mapillary Vistas Research Edition.
runtime n/a
subsampling no
submission date May, 2020
previous submissions 1, 2

 

Average results

Metric Value
AP 40.6385
AP50% 66.4262
AP100m 55.5747
AP50m 57.0377

 

Class results

Class AP AP50% AP100m AP50m
person 37.8683 69.5398 56.5293 56.5924
rider 32.1804 68.786 46.2341 46.8677
car 58.3448 81.6957 78.5634 80.7267
truck 44.1817 56.2493 59.6405 63.079
bus 53.4974 69.236 72.7039 79.4761
train 39.4568 60.7667 50.2899 48.5729
motorcycle 34.3716 66.237 43.6478 44.1895
bicycle 25.2067 58.8993 36.9885 36.7972

 

Links

Download results as .csv file

Benchmark page