Method Details


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

 

Method overview

name Panoptic-DeepLab [Mapillary Vistas]
challenge panoptic 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

 

Average results

Metric AllThingsStuff
PQ 66.477 58.82 72.0458
SQ 83.509 81.0524 85.2956
RQ 78.8491 72.5046 83.4632

 

Class results

Class PQ SQ RQ
road 98.6581 98.7232 99.9341
sidewalk 79.8022 86.6432 92.1043
building 90.0701 91.98 97.9236
wall 44.4291 79.7823 55.688
fence 46.7004 78.0795 59.8113
pole 68.8186 74.4253 92.4667
traffic light 59.8717 79.8638 74.9672
traffic sign 74.4994 83.1757 89.5687
vegetation 91.5944 92.3509 99.1809
terrain 47.4929 79.7943 59.5192
sky 90.5667 93.4336 96.9316
person 58.4592 78.1581 74.7961
rider 55.2847 74.2087 74.499
car 70.5909 85.3429 82.7144
truck 57.6905 87.6574 65.8135
bus 67.5217 88.5234 76.2755
train 57.3006 83.9403 68.2635
motorcycle 54.2332 77.4199 70.0508
bicycle 49.479 73.1681 67.6237

 

Links

Download results as .csv file

Benchmark page