Felidae | Cat Species Recognition Dataset

This dataset is meticulously compiled to support the identification and recognition of various cat species. It encompasses a broad assortment of images, each featuring a distinct species of cat.

Dataset Details

Image Characteristics: The images in this dataset are unnormalized to maintain the natural variations in appearance between specimens. Each image features a single cat without the presence of humans, other animals, or prey, ensuring that the focus remains solely on the cat species.
Selection Criteria:
The primary subject must be identifiable as one of the selected species.
Only images with a single specimen are included to avoid confusion in recognition tasks.

Diversity in Context: The dataset includes images of cats from diverse environments such as the wild, zoos, and conservation centers. It captures various perspectives, including rear views and unconventional angles, to enhance the robustness of recognition algorithms.
Additional Inclusions: Images of cubs and detailed fur close-ups are also part of this collection to provide comprehensive coverage of age-related and textural variations within each species.
Baseline Accuracy: The initial baseline model accuracy for this dataset is 0.2022, providing a benchmark for improvements in cat species recognition tasks.