This model classifies animal scat images (e.g., bear, deer, coyote, fox) using a fine-tuned Vision Transformer (ViT). Trained on ~3,000 images from a scat dataset.