Animations Dataset
nyuuzyou

folder main (5 files)
fileREADME.md 2.78kB
filetrain-00000.parquet 1.24GB
filetrain-00001.parquet 1.23GB
filetrain-00002.parquet 1.37GB
filetrain-00003.parquet 531.31MB
Type: Dataset
Tags:

Bibtex:
@article{,
title= {Animations Dataset},
journal= {},
author= {nyuuzyou},
year= {},
url= {https://huggingface.co/datasets/nyuuzyou/animations},
abstract= {# Dataset Card for Animations Dataset

### Dataset Summary
This dataset contains 50,849 animations with their associated metadata and source images. Each animation consists of multiple frames composed of simple sketch-level drawings, text elements, and potentially embedded images. The dataset provides complete information about each animation, including frame components, source images, timing between frames, and canvas settings. This makes it suitable for various tasks such as animation analysis, generation, and modification.

### Languages
The dataset is primarily monolingual:
- English (en): Any text elements within animations are predominantly in English.

## Dataset Structure

### Data Files
The dataset is stored as Parquet files with ZSTD compression:
- `train-00000.parquet` through `train-00003.parquet`
- Total: 4 shards, ~4.2 GB compressed

### Data Fields
Each row in the Parquet files contains the following columns:

| Column | Type | Description |
|--------|------|-------------|
| `id` | `string` | Unique identifier (UUID) for the animation |
| `settings` | `string` | JSON object containing canvas configuration |
| `dtimes` | `list[int64]` | Time delays between frames in milliseconds |
| `frames_data` | `string` | JSON array describing each frame's elements |
| `images` | `list[binary]` | PNG images used in the animation (decoded bytes) |

#### Settings Object
The `settings` JSON contains:
- `canvas_width`, `canvas_height`: Dimensions of the animation canvas
- `fillcolor`: Background color of the canvas (if specified)
- `default_font`: Default font used for text elements
- `default_font_size`: Default font size

#### Frames Data Structure
The `frames_data` JSON is an array of arrays, where each inner array represents a frame's elements:
- `type_for_loader`: Element type (e.g., "text", "image")
- `data`: Object containing element properties:
  - `type`: Element type
  - `centerx`, `centery`: Position coordinates on the canvas
  - `text`: Text content (for text elements)
  - `font`, `size`: Font properties
  - `rotate_angle`, `angle`: Rotation properties
  - `strokeColor`, `fillColor`, `textColor`: Color properties
  - `src`: Index into the `images` array (for image elements)
- `children_data`: Array of child elements (if any)

### Data Splits

| Split | Number of Examples |
|-------|--------------------|
| `train` | 50,849 |},
keywords= {},
terms= {},
license= {},
superseded= {}
}


Send Feedback