ANITA DATASET

An Industrial Animation Dataset

INTRODUCTION

Anita Dataset is a cartoon animation dataset contributed by animators who share licensed, free and professional data for academic research purposes. It contains authentic hand-drawings(i.e. tie-downs, colored sketches) from modern animation industry. We hope this open-source dataset facilitate cartoon research in providing more effective cartoon solutions. 

Thank you to all the animators who are willing to share your wonderful works. Without your contributions, this dataset would not be possible.

PUBLIC LICENSED REELS

CC BY-NC-SA 4.0 | CC BY

Image 1 Image 1 Hover

Polaris ©ONE_XT

Image 2 Image 2 Hover

Hope ©ONE_XT

Image 4 Image 4 Hover

Dogmatism @_嘲讽怪_

Image 5 Image 5 Hover

Embrace Freedom @日勿广隶

Image 6 Image 6 Hover

Zixu's 2023 Reels ©秄虚

DEMONSTRATION VIDEO

DATASET COMPOSITION

Anita Dataset mainly comprises three intermediate formats used in the modern animation industry: tie-down sketch, segment color, and composition. Image sequences are provided in 1080p resolution and are exported directly from engineering files, ensuring the quality of the data. Some animation works provide sketch-color pairs, enabling supervised learning tasks.

Image 1

Tie-down Sketch

Image 2

Segment Color

Image 3

Composition

Anita Dataset includes 16,000+ keyframes from 14 cartoons, covering 212 sketch cuts, 137 color cuts, and 18 compositions, showcasing various artistic styles and standards. The statistical composition of keyframes is shown below. For more information about each work, please refer to our GitHub repository of the project.

Image 1

DOWNLOAD

Anita Dataset is licensed under CC BY-NC-SA 4.0 and CC BY. Works from our dataset are licensed under either of these licenses. Please refer to the GitHub repository for more information. All copyrights are retained by the creators. By downloading this dataset, users agree to the terms specified in the licensing protocol.

If you find Anita Dataset useful, please consider citing our work.

@misc{Anita2024,
    title = {Anita Dataset},
    howpublished = {https://zhenglinpan.github.io/AnitaDataset_homepage/},
    note = {Accessed: 2024-06-24}
}

ACKNOWLEDGEMENTS

Anita Dataset is made possible by the animators(in alphabetical order):

The dataset is collected, organized, distributed by:

Contact

zhengli3@ualberta.ca