PyTorch Image Classification Inference

This function uses PyTorch for image classification, loads image data from a specified path, performs inference with a pre-trained model, and returns the predicted class.

Technology Stack : PyTorch, torchvision

Code Type : Image classification

Code Difficulty : Intermediate

                
                    
import random
import torch
import torchvision.transforms as transforms
import torchvision.datasets as datasets
from torch.utils.data import DataLoader

def random_image_classification(model, dataset_path, batch_size=32, num_workers=4):
    # Load the dataset
    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])
    ])
    dataset = datasets.ImageFolder(root=dataset_path, transform=transform)
    dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True, num_workers=num_workers)
    
    # Initialize the model
    model = model.to(device='cuda' if torch.cuda.is_available() else 'cpu')
    
    # Perform inference
    with torch.no_grad():
        for images, labels in dataloader:
            images = images.to(device)
            outputs = model(images)
            _, predicted = torch.max(outputs, 1)
            break

    return predicted