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).
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 |