
The author introduces the MAE-like framework to enhance nighttime images by utilizing severe augmentation during training to develop strong network priors resilient to real-world night haze effects.


coremsg

nighthaze-nighttime-image-dehazing-via-self-prior-learning


NightHaze: Nighttime Image Dehazing via Self-Prior Learning


title_rewrite


Severe augmentation during training yields strong network priors for effective nighttime image dehazing.



Proposing a semi-supervised model for real-world nighttime dehazing with spatial-frequency awareness and realistic brightness constraint.


a-comprehensive-study-on-semi-supervised-nighttime-dehazing-baseline-with-spatial-frequency-awareness-and-realistic-brightness-constraint


A Comprehensive Study on Semi-supervised Nighttime Dehazing Baseline with Spatial-Frequency Awareness and Realistic Brightness Constraint
