Clear2Fog generates realistic synthetic fog from clear scenes, enabling mixed-density training that outperforms full fixed-density data and improves real-world performance by 1.67 mAP after learning-rate adjustment.
Jarraud,Guide to Meteorological Instruments and Methods of Observation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Data Efficiency Study of Synthetic Fog for Object Detection Using the Clear2Fog Pipeline
Clear2Fog generates realistic synthetic fog from clear scenes, enabling mixed-density training that outperforms full fixed-density data and improves real-world performance by 1.67 mAP after learning-rate adjustment.