Method Details

Details for method 'MSeg1080_RVC'

name	MSeg1080_RVC
challenge	pixel-level semantic labeling
details	We present MSeg, a composite dataset that unifies semantic segmentation datasets from different domains. A naive merge of the constituent datasets yields poor performance due to inconsistent taxonomies and annotation practices. We reconcile the taxonomies and bring the pixel-level annotations into alignment by relabeling more than 220,000 object masks in more than 80,000 images, requiring more than 1.34 years of collective annotator effort. The resulting composite dataset enables training a single semantic segmentation model that functions effectively across domains and generalizes to datasets that were not seen during training. We adopt zero-shot cross-dataset transfer as a benchmark to systematically evaluate a model’s robustness and show that MSeg training yields substantially more robust models in comparison to training on individual datasets or naive mixing of datasets without the presented contributions.
publication	MSeg: A Composite Dataset for Multi-domain Semantic Segmentation John Lambert, Zhuang Liu, Ozan Sener, James Hays, Vladlen Koltun CVPR 2020 http://vladlen.info/papers/MSeg.pdf
project page / code	https://github.com/mseg-dataset/mseg-semantic
used Cityscapes data	fine annotations
used external data	COCO, ADE20K, SUN RGB-D, Mapillary Vistas, IDD, BDD, Cityscapes
runtime	0.49 s
subsampling	no
submission date	July, 2020
previous submissions