File size: 4,376 Bytes
f224ba4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bf5ebd3
 
f224ba4
 
 
 
bf5ebd3
f224ba4
 
bf5ebd3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f224ba4
 
 
 
bf5ebd3
 
 
 
 
 
f224ba4
bf5ebd3
f224ba4
 
bf5ebd3
f224ba4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
import torch
import torch.nn as nn
import torch.nn.functional as F
from torchvision import transforms
from PIL import Image
import numpy as np
import gradio as gr
import cv2
import matplotlib.pyplot as plt

# Define your CNN model
class TeethCNN(nn.Module):
    def __init__(self, num_classes=7):
        super(TeethCNN, self).__init__()
        self.conv1 = nn.Conv2d(3, 32, kernel_size=3, padding=1)
        self.conv2 = nn.Conv2d(32, 64, kernel_size=3, padding=1)
        self.conv3 = nn.Conv2d(64, 128, kernel_size=3, padding=1)
        self.conv4 = nn.Conv2d(128, 256, kernel_size=3, padding=1)
        self.pool = nn.MaxPool2d(2, 2)
        self.dropout = nn.Dropout(0.3)
        self.fc1 = nn.Linear(256 * 14 * 14, 256)
        self.fc2 = nn.Linear(256, num_classes)

    def forward(self, x):
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = self.pool(F.relu(self.conv3(x)))
        x = self.pool(F.relu(self.conv4(x)))
        x = x.view(x.size(0), -1)
        x = self.dropout(F.relu(self.fc1(x)))
        x = self.fc2(x)
        return x

# GradCAM logic
class GradCAM:
    def __init__(self, model, target_layer):
        self.model = model
        self.target_layer = target_layer
        self.gradients = None
        self.activations = None
        self._register_hooks()
        
    def _register_hooks(self):
        def forward_hook(module, input, output):
            self.activations = output

        def backward_hook(module, grad_input, grad_output):
            self.gradients = grad_output[0]

        self.target_layer.register_forward_hook(forward_hook)
        self.target_layer.register_full_backward_hook(backward_hook)

    def generate(self, input_tensor, class_idx=None):
        self.model.eval()
        output = self.model(input_tensor)
        if class_idx is None:
            class_idx = output.argmax(dim=1).item()
        loss = output[:, class_idx]
        self.model.zero_grad()
        loss.backward()
        gradients = self.gradients[0]
        activations = self.activations[0]
        weights = gradients.mean(dim=(1, 2))
        cam = torch.zeros(activations.shape[1:], device=activations.device)
        for i, w in enumerate(weights):
            cam += w * activations[i]
        cam = torch.relu(cam)
        cam = cam - cam.min()
        cam = cam / cam.max()
        return cam.detach().cpu().numpy()

# Load model
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
class_names = ['CaS', 'CoS', 'Gum', 'MC', 'OC', 'OLP', 'OT']
model = TeethCNN(num_classes=len(class_names))
model.load_state_dict(torch.load("teeth_model_weights.pth", map_location=device))
model.to(device)
model.eval()

# Preprocessing
transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize([0.5, 0.5, 0.5], 
                         [0.5, 0.5, 0.5])
])

def predict_with_gradcam(image):
    image = image.convert("RGB")
    input_tensor = transform(image).unsqueeze(0).to(device)

    # Prediction
    output = model(input_tensor)
    pred_idx = output.argmax(dim=1).item()
    pred_label = class_names[pred_idx]

    # Prepare base image
    img_np = np.array(image.resize((224, 224))) / 255.0

    # Multiple layer Grad-CAMs
    target_layers = [model.conv2, model.conv3, model.conv4]
    visualizations = []

    for layer in target_layers:
        gradcam = GradCAM(model, layer)
        cam = gradcam.generate(input_tensor)
        cam_resized = cv2.resize(cam, (224, 224))
        cam_overlay = cv2.applyColorMap(np.uint8(255 * cam_resized), cv2.COLORMAP_JET)
        cam_overlay = cv2.cvtColor(cam_overlay, cv2.COLOR_BGR2RGB)
        overlay = (0.5 * img_np + 0.5 * cam_overlay / 255.0)
        overlay = np.clip(overlay, 0, 1)
        visualizations.append(overlay)

    return pred_label, *visualizations


interface = gr.Interface(
    fn=predict_with_gradcam,
    inputs=gr.Image(type="pil"),
    outputs=[
        gr.Label(label="Predicted Class"),
        gr.Image(label="Grad-CAM: Conv2"),
        gr.Image(label="Grad-CAM: Conv3"),
        gr.Image(label="Grad-CAM: Conv4")
    ],
    title="🦷 Teeth Disease Classifier with Grad-CAM",
    description="Upload a teeth image. The model predicts the class and shows Grad-CAM visualizations for multiple convolutional layers."
)


interface.launch()