lsqkk/animal-recognition-dataset

Animal Recognition Dataset - 动物识别数据集

EN / 中文

A high-quality animal image dataset for computer vision research and CNN model training. This dataset contains 9,757 carefully curated animal images across multiple categories, suitable for object recognition, classification, and transfer learning tasks.

一个用于计算机视觉研究和CNN模型训练的高质量动物图像数据集。本数据集包含9,757张经过精心筛选的动物图像，涵盖多个类别，适用于目标识别、分类和迁移学习任务。

Dataset Overview

• Total Images: 9,757
• Categories: 16 distinct animal categories
• Image Format: JPG
• Image Size: All images are under 80KB while maintaining high visual quality
• Resolution: Various resolutions, optimized for training efficiency
• Dataset Size: 326 MB

Directory Structure

.
├── cat/
│   ├── cat_mixed/        # 579 images
│   ├── jiafei/           # 255 images (加菲猫)
│   ├── jumao/            # 772 images (橘猫)
│   └── sanhua/           # 84 images (三花猫)
├── dog/
│   ├── dog_mixed/        # 995 images
│   ├── fadou/            # 590 images (法斗)
│   ├── hashiqi/          # 890 images (哈士奇)
│   ├── jinmao/           # 634 images (金毛)
│   ├── keji/             # 682 images (柯基)
│   └── samoye/           # 1,139 images (萨摩耶)
├── livestock/
│   ├── cattle/           # 369 images (牛)
│   └── horse/            # 224 images (马)
├── man/
│   ├── blackman/         # 798 images
│   └── whiteman/         # 745 images
└── poultry/
    ├── chicken/          # 367 images (鸡)
    └── goose/            # 634 images (鹅)

Category Distribution

Category	Images	Percentage
Samoye	1,139	11.67%
Dog (Mixed)	995	10.20%
Husky	890	9.12%
Black Man	798	8.18%
Jumao Cat	772	7.91%
White Man	745	7.64%
Corgi	682	6.99%
Golden Retriever	634	6.50%
Goose	634	6.50%
French Bulldog	590	6.05%
Cat (Mixed)	579	5.93%
Cattle	369	3.78%
Chicken	367	3.76%
Persian Cat	255	2.61%
Horse	224	2.30%
Sanhua Cat	84	0.86%

Dataset Features

1. Diverse Categories: Covering pets, livestock, poultry, and human categories for comprehensive recognition tasks
2. Balanced Distribution: While not perfectly balanced, the dataset provides sufficient samples for each category
3. High Quality: All images are carefully selected for clarity and relevance
4. Standardized Format: Consistent JPG format with size optimization
5. Hierarchical Organization: Logical folder structure for easy data loading

Intended Use Cases

• Image classification model training
• Object detection and recognition
• Transfer learning experiments
• CNN architecture benchmarking
• Computer vision course projects
• Animal recognition system development

Usage Example

from torchvision import datasets, transforms

# Data loading example
transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

dataset = datasets.ImageFolder(root='./path/to/dataset', transform=transform)

Dataset Splits Recommendation

For training/validation/test splits, we recommend:

• Training: 70% of images
• Validation: 15% of images
• Test: 15% of images

Maintain class distribution ratios when splitting to ensure balanced evaluation.

Future Expansion

This dataset is actively maintained and expanded. Planned additions include:

• More domestic animal breeds
• Wild animal categories
• Exotic bird species
• Marine animals
• Additional livestock varieties

Contributions and suggestions for new categories are welcome.

Ethical Considerations

The dataset includes human images categorized by apparent skin tone. Researchers should:

1. Use these categories responsibly and avoid reinforcing biases
2. Consider the ethical implications of skin tone classification
3. Apply appropriate fairness metrics during model evaluation
4. Document limitations regarding demographic representation

License and Usage

This dataset is released under the MIT License.

Important Notice:

• All images are collected from publicly available sources on the internet
• This dataset is intended strictly for non-commercial research and educational purposes
• Commercial use is strictly prohibited
• If you are the copyright owner of any image in this dataset and wish to have it removed, please contact us immediately
• We respect intellectual property rights and will promptly remove any infringing material upon notification

Copyright Notice:

If you believe your copyrighted work appears in this dataset without authorization, please contact us at: jsxzznz@163.com

We will investigate and remove the infringing content within 48 hours of notification.

Citation

If you use this dataset in your research, please cite it as:

lsqkk. (2024). Animal Recognition Dataset [Data set]. GitHub. 
https://github.com/lsqkk/animal-recognition-dataset

Contributing

Contributions to expand and improve this dataset are welcome. Please:

1. Ensure images are under 80KB
2. Maintain JPG format
3. Follow the existing directory structure
4. Include diverse and high-quality images
5. Submit pull requests with clear descriptions

Contact

For dataset-related inquiries, suggestions, or copyright concerns:

• Email: jsxzznz@163.com
• GitHub Issues: Create an issue

Acknowledgments

We thank the open-source community and all contributors who have helped in curating this dataset for research purposes.

---

Note: This dataset is provided "as-is" without any warranty. Users are responsible for ensuring their use complies with applicable laws and ethical guidelines.