image
imagewidth (px)
1.02k
1.02k
label
int64
0
0
filename
stringlengths
5
9
0
1.png
0
10.png
0
100.png
0
10000.png
0
10003.png
0
10007.png
0
10009.png
0
10014.png
0
10016.png
0
10023.png
0
10024.png
0
1003.png
0
10031.png
0
1004.png
0
10040.png
0
10048.png
0
1005.png
0
10053.png
0
10054.png
0
10056.png
0
10058.png
0
10061.png
0
10063.png
0
10065.png
0
10066.png
0
10069.png
0
10074.png
0
10076.png
0
10078.png
0
10080.png
0
10086.png
0
10087.png
0
10088.png
0
10089.png
0
1009.png
0
10091.png
0
10094.png
0
1010.png
0
10101.png
0
10108.png
0
10110.png
0
10115.png
0
10116.png
0
1012.png
0
10123.png
0
10125.png
0
10128.png
0
1013.png
0
10130.png
0
10131.png
0
10134.png
0
10136.png
0
10139.png
0
10142.png
0
10143.png
0
10146.png
0
10148.png
0
1015.png
0
10150.png
0
10153.png
0
10155.png
0
10156.png
0
10158.png
0
1016.png
0
10163.png
0
10164.png
0
10166.png
0
10170.png
0
10171.png
0
10179.png
0
10180.png
0
10181.png
0
10187.png
0
10188.png
0
1019.png
0
10190.png
0
10192.png
0
10193.png
0
10195.png
0
102.png
0
10201.png
0
10203.png
0
10205.png
0
10206.png
0
10209.png
0
10210.png
0
10212.png
0
10213.png
0
10220.png
0
10221.png
0
10222.png
0
10225.png
0
10226.png
0
10231.png
0
10232.png
0
10236.png
0
1024.png
0
10240.png
0
10241.png
0
10244.png

ShinkaiGAN Image Dataset

File structure

├── train/ 
│   ├── real/ 
│   ├── anime/ 
│   └── shinkai/ 
├── val/ 
│   ├── real/ 
│   ├── anime/ 
│   └── shinkai/ 
└── test/ 
    ├── real/ 
    ├── anime/ 
    └── shinkai/

How to access it:

Dataset["{split}"][{image index}]["{feature}"]
  • split: train, test, validation
  • image index: pick any interger from the amount*
  • features: label, filename, image**

*if you want to pick train anime image then pick from 0-23999, if you want to pick train real images then 24000-47999, and so on.
**label 0 is generated anime, label 1 is real images, label 2 is makoto shinkai film
**filename is the name of the file
**image is the feature containing the images it's in PIL <PIL.PngImagePlugin.PngImageFile image mode=RGB size=1024x1024>

example:

dataset["train"][0]["image"]

Dataset Overview

Pair Dataset

  • Real
  • Anime

Unpair Dataset

  • Shinkai

Image Amount

  • Train: 24,000 images x 3 = 72000 images
  • Validation: 5,000 images x 3 = 15000 images
  • Test: 1,000 images x 3 = 3000 images

Image Composition

  • Real: LHQ1024

  • Anime:

    • Generated from the real dataset using Stable Diffusion with the following prompts:
      • Prompt: "anime artwork anime style, landscape. anime style, key visual, vibrant, studio anime, highly detailed, in makoto shinkai style"
      • Negative Prompt: "photo, deformed, black and white, realism, disfigured, low contrast, cityscape, people, meteor, sunset."
      • Parameters:
        • num_steps = 10
        • guidance_scale = 0.2
        • controlnet_conditioning_scale = 0.9
        • seed = random.randint(0000000000, 9999999999)
  • Shinkai:

    • Sourced from Makoto Shinkai anime in 1080p, cropped to 1080 x 1080.
    • Titles Included:
      • Your Name (Kimi no nawa)
      • Weathering with You (Tenki no Ko)
      • Suzume
      • Garden of Words
      • 5 Centimeters per Second
      • Children Who Chase Lost Voices

Contributor

  • Dwiyasa Nakula - Sepuluh Nopember Institute of Technology (Institut Technologi Sepuluh Nopember - ITS)
  • ShinkaiGAN Team - Sepuluh Nopember Institute of Technology (Institut Technologi Sepuluh Nopember - ITS)
Downloads last month
54