A non-adversarial pyramid curriculum that unifies self-training pseudo-labels with curriculum label-distribution constraints achieves state-of-the-art segment adaptation from synthetic to real city images.
Curriculum model adaptation with synthetic and real data for semantic foggy scene understanding
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2019 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach
A non-adversarial pyramid curriculum that unifies self-training pseudo-labels with curriculum label-distribution constraints achieves state-of-the-art segment adaptation from synthetic to real city images.