Skip to content

Latest commit

 

History

History
14 lines (8 loc) · 812 Bytes

README.md

File metadata and controls

14 lines (8 loc) · 812 Bytes

Unified Perceptual Parsing for SceneUnderstanding

Reference

Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, Jian Sun. "Unified Perceptual Parsing for Scene Understanding." Proceedings of the European Conference on Computer Vision (ECCV). 2018.

Performance

Cityscapes

Model Backbone Resolution Training Iters mIoU mIoU (flip) mIoU (ms+flip) Links
UPerNet ResNet101_OS8 512x1024 40000 79.58% 80.11% 80.41% model|log|vdl