6 papers · 1 filter
Degradation-Consistent Learning via Bidirectional Diffusion for Low-Light Image Enhancement
Jinhong He, Minglong Xue, Zhipu Liu +3
Low-light image enhancement aims to improve the visibility of degraded images to better align with human visual perception. While diffusion-based methods have shown promising perfo…
DFDNet: Dynamic Frequency-Guided De-Flare Network
Minglong Xue, Aoxiang Ning, Shivakumara Palaiahnakote +1
Strong light sources in nighttime photography frequently produce flares in images, significantly degrading visual quality and impacting the performance of downstream tasks. While s…
KAN See In the Dark
Aoxiang Ning, Minglong Xue, Jinhong He +1
Existing low-light image enhancement methods are difficult to fit the complex nonlinear relationship between normal and low-light images due to uneven illumination and noise effect…
Zero-Shot Low-Light Image Enhancement via Joint Frequency Domain Priors Guided Diffusion
Jinhong He, Shivakumara Palaiahnakote, Aoxiang Ning +1
Due to the singularity of real-world paired datasets and the complexity of low-light environments, this leads to supervised methods lacking a degree of scene generalisation. Meanwh…
Addressing Domain Discrepancy: A Dual-branch Collaborative Model to Unsupervised Dehazing
Shuaibin Fan, Minglong Xue, Aoxiang Ning +1
Although synthetic data can alleviate acquisition challenges in image dehazing tasks, it also introduces the problem of domain bias when dealing with small-scale data. This paper p…
Artistic-style text detector and a new Movie-Poster dataset
Aoxiang Ning, Yiting Wei, Minglong Xue +1
Although current text detection algorithms demonstrate effectiveness in general scenarios, their performance declines when confronted with artistic-style text featuring complex str…