OpenRR-1k is a 1,000-pair real-world reflection removal dataset generated by off-the-shelf AI removal plus manual refinement, with benchmarks showing fine-tuning gains.
Deep learning based end-to-end specular reflection re- moval for medical endoscopic images,
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
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
OpenRR-1k: A Scalable Dataset for Real-World Reflection Removal
OpenRR-1k is a 1,000-pair real-world reflection removal dataset generated by off-the-shelf AI removal plus manual refinement, with benchmarks showing fine-tuning gains.