CRWKV, a bidirectional RWKV network with context-guided token shift and frequency mixing, matches or modestly exceeds prior real-image denoisers with lower computational cost.
Efficientvit: Lightweight multi-scale attention for high-resolution dense prediction
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Multi-View Learning with Context-Guided Receptance for Image Denoising
CRWKV, a bidirectional RWKV network with context-guided token shift and frequency mixing, matches or modestly exceeds prior real-image denoisers with lower computational cost.