Method Details


Details for method 'iFLYTEK-CV'

 

Method overview

name iFLYTEK-CV
challenge panoptic semantic labeling
details iFLYTEK Research, CV Group
publication Anonymous
project page / code
used Cityscapes data fine annotations, coarse annotations
used external data coco态mapillary
runtime n/a
subsampling no
submission date November, 2019
previous submissions

 

Average results

Metric AllThingsStuff
PQ 65.9657 57.9858 71.7692
SQ 83.1951 81.2811 84.5871
RQ 78.5027 71.2761 83.7585

 

Class results

Class PQ SQ RQ
road 98.7247 98.8224 99.9011
sidewalk 79.5644 86.5903 91.8861
building 90.1463 91.9628 98.0248
wall 45.5432 78.6043 57.9399
fence 46.2984 76.3415 60.6464
pole 65.4337 72.8418 89.8299
traffic light 59.6248 78.15 76.2953
traffic sign 74.4659 81.972 90.843
vegetation 91.5271 92.4083 99.0463
terrain 47.8509 79.2814 60.3558
sky 90.2813 93.4835 96.5746
person 59.192 78.4469 75.4548
rider 56.5796 75.5868 74.8538
car 69.9771 85.3043 82.0323
truck 54.9372 88.4089 62.1399
bus 63.7231 88.0953 72.3343
train 61.4888 83.9559 73.2394
motorcycle 51.8325 77.0263 67.2919
bicycle 46.1564 73.4247 62.8623

 

Links

Download results as .csv file

Benchmark page