You can deploy this model and start earning money today!

ujjwal75/
indian-bovine-breeds-model

Name: indian-bovine-breeds-model
Availability: InStock

Views

No views yet

🐄 Indian Bovine Breeds Model

A deep learning model for classifying Indian bovine (cow and buffalo) breeds from images.
It is fine-tuned to recognize 40 popular Indian cattle and buffalo breeds using convolutional neural networks (CNNs).

📘 Overview

This model helps identify bovine breeds from uploaded images.
It can be useful in livestock research, breed conservation, farm management, and educational tools.

Model name: ujjwal75/indian-bovine-breeds-model
File size: 334 MB
Framework: PyTorch
Author: @ujjwal75
Supported breeds: 40 (see list below)

🧠 Supported Breeds

Alambadi, Amritmahal, Ayrshire, Banni, Bargur, Bhadawari, Brown_Swiss, Dangi, Deoni, Gir, Guernsey, Hallikar, Hariana, Holstein_Friesian, Jaffrabadi, Jersey, Kangayam, Kankrej, Kasargod, Kenkatha, Kherigarh, Khillari, Krishna_Valley, Malnad_gidda, Mehsana, Murrah, Nagori, Nagpuri, Nili_Ravi, Nimari, Ongole, Pulikulam, Rathi, Red_Dane, Red_Sindhi, Sahiwal, Surti, Tharparkar, Toda, Umblachery, Vechur

📄 Class mapping file: classes.json

⚙️ Model Details

Property	Description
Architecture	Fine-tuned CNN (ResNet-50 / EfficientNet-B0)
Framework	PyTorch
Input	RGB image (JPEG/PNG)
Output	Predicted breed label + confidence score
Loss Function	CrossEntropyLoss
Optimizer	Adam / SGD
Epochs	~30
Image Size	224×224
Accuracy	~90% on validation data

🧩 Dataset & Training

Dataset: Custom-curated collection of Indian bovine images (40 breeds)
Split: 80% training / 20% validation
Preprocessing: Resize to 224×224, normalization using ImageNet mean and std
Augmentation: Random rotation, horizontal flip, and brightness adjustment
Hardware: Trained on GPU (NVIDIA CUDA)
Model File: Indian_bovine_finetuned_model.pth (334 MB)

🚀 Usage

🔹 Python Example

python

from PIL import Image
import requests
import torch
from torchvision import transforms
from huggingface_hub import hf_hub_download
import json
import torchvision.models as models

# Download model and class labels
model_path = hf_hub_download("ujjwal75/indian-bovine-breeds-model", "Indian_bovine_finetuned_model.pth")
classes_path = hf_hub_download("ujjwal75/indian-bovine-breeds-model", "classes.json")

# Load classes
with open(classes_path, "r") as f:
    classes = json.load(f)

# Load pretrained architecture (adjust if different)
model = models.resnet50(num_classes=len(classes))
model.load_state_dict(torch.load(model_path, map_location="cpu"))
model.eval()

# Image preprocessing
url = "https://example.com/bovine.jpg"
image = Image.open(requests.get(url, stream=True).raw)
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]),
])
input_tensor = transform(image).unsqueeze(0)

# Prediction
with torch.no_grad():
    outputs = model(input_tensor)
    pred_idx = torch.argmax(outputs, dim=1).item()

print("Predicted Breed:", classes[pred_idx])