Mammals Image Classification Dataset - 45 Animals
This dataset contains 13,751 mammal images divided into 45 distinct classes. Each class is stored in a specific folder, following the classic ImageNet structure.
13,751 JPG images, 256x256 pixel resolution, 45 classes
ODC Attribution License (ODC-by)
Description
Mammals Image Classification Dataset - 45 Animals includes images of mammalian animals classified into 45 categories. The images are in JPG format with a standardized resolution of 256x256 pixels, ideal for image classification and deep learning tasks.
What is this dataset for?
- Train animal image classification models
- Develop automatic cash recognition systems
- Testing CNN architectures on various animal images
Can it be enriched or improved?
This dataset can be enriched by additional annotation (e.g.: location of animals in the image), or by increasing the data (transformations, filters) to improve the robustness of the models.
🔎 In summary
🧠 Recommended for
- Computer vision researchers
- Students
- AI developers
🔧 Compatible tools
- TensorFlow
- PyTorch
- Keras
- FastAI
💡 Tip
Use image augmentation to improve robustness in the face of variations in lighting and angles.
Frequently Asked Questions
How many animal classes are there in this dataset?
There are 45 distinct classes representing different mammal species.
Are the images annotated for the location of the animals?
No, only image classification labels, no location annotations.
Is this dataset suitable for training computer vision models for beginners?
Yes, its clear organization and volume make it a great choice for beginners in image classification.
