---
tags:
- jigglypuff
- pokemon
- kirby
pretty_name: ' Jigglypuff detection data'
size_categories:
- 10K<n<100K
---

# Jigglypuff Detection Dataset

This dataset is designed for training machine learning models to detect and distinguish Jigglypuff (a Pokemon) from Kirby and other similar characters.

I mainly use this with my [Pummeluff-Detektor](https://github.com/PlexSheep/pummeluff_detektor)

## Dataset Summary

The dataset contains approximately 5.5GB of images across three categories:
- Jigglypuff (Pokemon character)
- Kirby (Nintendo character)
- Other (Various other characters and images)

This dataset was specifically curated to solve the age-old problem of distinguishing between these two famously similar pink characters, along with detecting potential imposters.

## Supported Tasks

- Image Classification
- Character Recognition
- Pink Blob Detection

## Dataset Structure

```
dataset/
├── jigglypuff/    (~730MB)
├── kirby/         (~4MB)
└── other/         (~4.8GB)
    ├── other/     
    └── other2/    
```

### Data Fields

The dataset is organized in a directory structure where each subdirectory represents a class label:
- `jigglypuff`: Contains Jigglypuff images
- `kirby`: Contains Kirby images, relevant because kirby and jigglypuff look similar
- `other`: Contains images of other characters and objects for negative examples

### Data Splits

This is a raw image dataset without predefined splits. Users should implement their own train/test splits.

## Dataset Creation

### Source Data

Images were collected from two primary sources:
- [CS63 Pokemon Dataset](https://www.kaggle.com/datasets/alinapalacios/cs63-pokemon-dataset)
- [Nintendo Kirby Dataset](https://www.kaggle.com/datasets/turbihao/nintendo-kirby-dataset)

As of writing this, no additional images were added.

### Annotations

The dataset uses directory names as implicit labels. No additional annotation files are provided.

## Considerations for Using the Dataset

### Social Impact of Dataset

This dataset was created for educational and entertainment purposes. It demonstrates practical machine learning concepts through a fun and accessible example.

### Discussion of Biases

The dataset may contain biases in terms of:
- Art styles (anime/game artwork vs. fan art)
- Image quality variations
- Pose distributions
- Background variations
- Only images of Pokemon or Kirby, no unrelated images

### Other Known Limitations

- Limited to specific art styles
- May not generalize well to real-world pink objects
- Focused on frontal/profile views

## Additional Information

### Dataset Curators

Dataset curated by @plexsheep

### Contributions

Feel free to open a discussion on the dataset repository page if you have:
- Suggestions for additional images
- Quality improvements
- Documentation updates
- Use case examples