UniSER is a unified diffusion transformer foundation model that removes diverse soft image degradations by training on a large curated dataset of semi-transparent occlusions with fine-grained controls.
Single image reflection removal via inter-layer complementarity
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
UniSER: A Foundation Model for Unified Soft Effects Removal
UniSER is a unified diffusion transformer foundation model that removes diverse soft image degradations by training on a large curated dataset of semi-transparent occlusions with fine-grained controls.