DiNAT-IR, a transformer that alternates local and dilated neighborhood attention and adds a channel-aware module, matches or exceeds state-of-the-art image restoration on several benchmarks, particularly motion deblurring.
Digital image restoration
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DiNAT-IR: Exploring Dilated Neighborhood Attention for High-Quality Image Restoration
DiNAT-IR, a transformer that alternates local and dilated neighborhood attention and adds a channel-aware module, matches or exceeds state-of-the-art image restoration on several benchmarks, particularly motion deblurring.