Skip to content

Latest commit

 

History

History
13 lines (8 loc) · 815 Bytes

README.md

File metadata and controls

13 lines (8 loc) · 815 Bytes

Deep dual-resolution networks for real-time and accurate semantic segmentation of road scenes

Reference

Yuanduo Hong, Huihui Pan, Weichao Sun, Yisong Jia. "Deep dual-resolution networks for real-time and accurate semantic segmentation of road scenes." arXiv preprint arXiv:2101.06085 (2021).

Performance

Cityscapes

Model Backbone Resolution Training Iters mIoU mIoU (flip) mIoU (ms+flip) Links
DDRNet_23 - 1024x1024 120000 79.85% 80.11% 80.59% model|log|vdl