Share: Link

1 Paper with Code

1.1 Browse State-of-the-Art

Link:

• https://paperswithcode.com/sota

CV:

• https://paperswithcode.com/area/computer-vision

Preview:

2 Datasets of CV

2.1 Machine learning datasets

Link:

• https://paperswithcode.com/datasets

Preview:

2.2 Data (Visual Geometry Group)

Link:

• HOME: https://www.robots.ox.ac.uk/~vgg/
• DATA: https://www.robots.ox.ac.uk/~vgg/data/
• Web DEMO: https://www.robots.ox.ac.uk/~vgg/demo/

Preview: (Data)

2.3 Datasets of Image Inpainting

2.3.1 Mask: NVIDIA Irregular Mask Dataset: Testing Set

Link:
https://nv-adlr.github.io/publication/partialconv-inpainting

大佬整理：
图像修复 mask 下载 | NVIDIA Irregular Mask Dataset | Quick Draw Irregular Mask Dataset | 墨理三生 |

2.3.2 Paris Street View Dataset

Link:
https://www.robots.ox.ac.uk/~vgg/data/parisbuildings/

A dataset containing 14,900 training images and 100 test images collected from street views of Paris. This dataset mainly focuses on the buildings in the city.

2.3.3 Places2 Challenge Dataset

Link:
http://places2.csail.mit.edu/download.html

A dataset released by MIT containing over 8,000,000 images from over 365 scenes, which is very suitable for building inpainting models as it enables the model to learn the distribution from many natural scenes.

2.3.4 CelebA Dataset

CelebA Dataset: Large-scale CelebFaces Attributes (CelebA) Dataset

Link:
http://mmlab.ie.cuhk.edu.hk/projects/CelebA.html

A dataset focusing on human face images and containing over 180,000 training images. The model trained on this dataset can easily be transferred to face editing/completion tasks.