Category ""
ADVERTISING
The dataset contains 4279 real-world underwater images, and the corresponding clear images are generated as comparison references. We also provide the semantic segmentation map and medium transmission map for each image. You can download it from here (lsui).
LSUI dataset is a large-scale underwater image dataset that has 5004 image pairs
The LSUI is a large-scale underwater image (LSUI) dataset, which involves richer underwater scenes (lighting conditions, water types and target categories) and better visual quality reference images than the existing ones.
LSUI stands for Large Scale Underwater Image Dataset, which is a large-scale underwater image dataset containing 5,004 image pairs. This dataset involves richer underwater scenes (lighting conditions, water types, and target categories) and has better visual quality than existing reference images.
To this end, this article proposes a new UIE network, underwater image multikernel convolution transformer network (UMCTN), based on a multikernel convolution Transformer. A multikernel convolutional residual attention (MCRA) was constructed.
我们发布了一个包含 5004 个图像对的大规模水下图像 (LSUI) 数据集，其中涉及比现有图像更丰富的水下场景（光照条件、水体类型和目标类别）和更好的视觉质量参考图像。
The json representation of the dataset with its distributions based on DCAT. Zheng Cheng, Guodong Fan, Jingchun Zhou, Min Gan, C. L. Philip Chen (2024). Dataset: LSUI. https://doi.org/10.57702/bnowj51m.
images are generated as comparison references. Furthermore, with the prior knowledge that the attenuation of different color channels and space areas in underwater images is inconsistent, we designed a channel-wise multi-scale feature fusion transformer (CMSFFT) and a spatial-wise global feature modeling transformer (SGFMT) based on the ...
In this work, we constructed a large-scale underwater image (LSUI) dataset including 4279 image pairs, and reported an U-shape Transformer network where the transformer model is for the first time introduced to the UIE task.
The LSUI is a large-scale underwater image (LSUI) dataset, which involves richer underwater scenes (lighting conditions, water types and target categories) and better visual quality reference images than the existing ones.