Spaces:
Sleeping
Sleeping
| ### 1. Imports and class names setup ### | |
| import gradio as gr | |
| import os | |
| import torch | |
| import numpy as np | |
| from model import create_gadgets_model | |
| from timeit import default_timer as timer | |
| from typing import Tuple, Dict | |
| # Setup class names | |
| class_names = ["Headphones","Laptop","Mobile"] | |
| ### 2. Model and transforms preparation ### | |
| # Create Gadget model | |
| gadget, gadget_transforms = create_gadgets_model( | |
| num_classes=3, # len(class_names) would also work | |
| ) | |
| # Load saved weights | |
| gadget.load_state_dict( | |
| torch.load( | |
| f="Gadgets_model_save.pth", | |
| map_location=torch.device("cpu"), # load to CPU | |
| ) | |
| ) | |
| ### 3. Predict function ### | |
| # Create predict function | |
| def predict(img) -> Tuple[Dict, float]: | |
| """Transforms and performs a prediction on img and returns prediction and time taken. | |
| """ | |
| # Start the timer | |
| start_time = timer() | |
| threshold=0.87 | |
| # Transform the target image and add a batch dimension | |
| img = gadget_transforms(img).unsqueeze(0) | |
| # Put model into evaluation mode and turn on inference mode | |
| gadget.eval() | |
| with torch.inference_mode(): | |
| # Pass the transformed image through the model and turn the prediction logits into prediction probabilities | |
| pred_probs = torch.softmax(gadget(img), dim=1) | |
| # Create a prediction label and prediction probability dictionary for each prediction class (this is the required format for Gradio's output parameter) | |
| y=(pred_probs>threshold).float() | |
| pred_prob=y.long() | |
| pred_prob=pred_prob.numpy() | |
| x=np.count_nonzero(pred_prob) | |
| if x==0: | |
| pred_labels_and_probs={"Unknown Images found...Please Provide Images of Headphone,Mobile or laptop":0} | |
| else: | |
| pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))} | |
| # Calculate the prediction time | |
| pred_time = round(timer() - start_time, 5) | |
| # Return the prediction dictionary and prediction time | |
| return pred_labels_and_probs, pred_time | |
| ### 4. Gradio app ### | |
| # Create title, description and article strings | |
| title = "Gadgets Classifier 📱🖥🎧" | |
| description = "An computer vision model to classify images of Gadgets as Headphone, Laptop and Mobile." | |
| article = "Created by Vishal Jadhav (www.linkedin.com/in/vishaljadhav1855)" | |
| # Create the Gradio demo | |
| demo = gr.Interface(fn=predict, # mapping function from input to output | |
| inputs=gr.Image(type="pil"), # what are the inputs? | |
| outputs=[gr.Label(num_top_classes=3, label="Predictions"), # what are the outputs? | |
| gr.Number(label="Prediction time (s)")], # our fn has two outputs, therefore we have two outputs | |
| examples=["Headphones17.jpg","Image_16.jpg","Image_20.jpg"], | |
| title=title, | |
| description=description, | |
| article=article) | |
| # Launch the demo! | |
| demo.launch() | |