Low-light Image Enhancement using Cross Attention

1. Model Architecture

• Below are some examples of the model's architecture and results:

  
  
  
  
  

---

2. Traffic-297 Dataset

• Download the training dataset:
  Traffic-297 Dataset
  Password: amk2

---

3. Model Weights and Results

Below are the pre-trained weights and corresponding results for various datasets:

Dataset	Checkpoints	Results
LOL-v1	Download (Password: 6rf7)	Result (Password: bqrb)
LOL-v2-r	Download (Password: vtqj)	Result (Password: eqmb)
LOL-v2-s	Download (Password: ax7q)	Result (Password: 5he3)
SID	Download (Password: hznl)	Result (Password: u4un)
SMID	Download (Password: ittd)	Result (Password: h9wx)
SDSD-in	Download (Password: 71zq)	Result (Password: e5ms)
SDSD-out	Download (Password: 7n3e)	Result (Password: bsed)
Traffic-297	Download (Password: mspy)	Result (Password: z2px)

---

4. Training

• The model is trained using the BasicSR toolbox.

Citation:

@misc{basicsr,
  author =       {Xintao Wang and Liangbin Xie and Ke Yu and Kelvin C.K. Chan and Chen Change Loy and Chao Dong},
  title =        {{BasicSR}: Open Source Image and Video Restoration Toolbox},
  howpublished = {\url{https://github.com/XPixelGroup/BasicSR}},
  year =         {2022}
}