PussyCat-vs-Doggie-SigLIP2

PussyCat-vs-Doggie-SigLIP2 is an image classification vision-language encoder model fine-tuned from google/siglip2-base-patch16-224 for a single-label classification task. It is designed to classify images as either a cat or a dog using the SiglipForImageClassification architecture.

The model categorizes images into two classes:

• Class 0: "Pussy Cat"
• Class 1: "Doggie"

Classification Report:
              precision    recall  f1-score   support

   Pussy Cat     0.9194    0.8745    0.8964     12500
      Doggie     0.8803    0.9234    0.9013     12500

    accuracy                         0.8989     25000
   macro avg     0.8999    0.8989    0.8989     25000
weighted avg     0.8999    0.8989    0.8989     25000

Run with Transformers🤗

!pip install -q transformers torch pillow gradio

import gradio as gr
from transformers import AutoImageProcessor
from transformers import SiglipForImageClassification
from transformers.image_utils import load_image
from PIL import Image
import torch

# Load model and processor
model_name = "prithivMLmods/PussyCat-vs-Doggie-SigLIP2"
model = SiglipForImageClassification.from_pretrained(model_name)
processor = AutoImageProcessor.from_pretrained(model_name)

def animal_classification(image):
    """Predicts whether the image contains a cat or a dog."""
    image = Image.fromarray(image).convert("RGB")
    inputs = processor(images=image, return_tensors="pt")
    
    with torch.no_grad():
        outputs = model(**inputs)
        logits = outputs.logits
        probs = torch.nn.functional.softmax(logits, dim=1).squeeze().tolist()
    
    labels = {
        "0": "Pussy Cat", 
        "1": "Doggie"
    }
    predictions = {labels[str(i)]: round(probs[i], 3) for i in range(len(probs))}
    
    return predictions

# Create Gradio interface
iface = gr.Interface(
    fn=animal_classification,
    inputs=gr.Image(type="numpy"),
    outputs=gr.Label(label="Prediction Scores"),
    title="Cat vs Dog Classification",
    description="Upload an image to classify whether it contains a cat or a dog."
)

# Launch the app
if __name__ == "__main__":
    iface.launch()

Intended Use:

The PussyCat-vs-Doggie-SigLIP2 model is designed to classify images as either a cat or a dog. Potential use cases include:

• Pet Identification: Helping users distinguish between cats and dogs.
• Automated Pet Sorting: Useful for shelters and pet adoption platforms.
• Educational Purposes: Assisting in teaching image classification concepts.
• Surveillance & Security: Identifying animals in security footage.