File size: 2,646 Bytes
d31170d
 
 
 
 
 
d82f8ec
 
d31170d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
473b629
d31170d
8f49a5b
d31170d
 
 
 
 
 
 
 
 
 
 
 
 
 
e5c1737
d31170d
59ac305
09898e2
b8cbc9e
d31170d
 
099df44
d31170d
 
 
 
 
 
d0c5619
d31170d
 
 
 
 
 
e5c1737
d31170d
52c6c75
d31170d
 
55ede45
 
d31170d
 
 
f7aadc0
 
 
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
import torch
import torchvision
from torchvision import transforms
import numpy as np
import gradio as gr
from PIL import Image
from pytorch_grad_cam import GradCAM
from pytorch_grad_cam.utils.image import show_cam_on_image
from resnet_lightning import ResNet18Model
import gradio as gr

model = ResNet18Model.load_from_checkpoint("epoch=19-step=3920.ckpt")

inv_normalize = transforms.Normalize(
    mean = [-0.50/0.23, -0.50/0.23, -0.50/0.23],
    std= [1/0.23, 1/0.23,1/0.23]
)

classes = ('plane', 'car', 'bird', 'cat', 'deer',
           'dog', 'frog', 'horse', 'ship', 'truck')

def resize_image_pil(image, new_width, new_height):
    img =  Image.fromarray(np.array(image))
    width, height = img.size

    width_scale = new_width/width
    height_scale = new_height/height
    scale = min(width_scale, height_scale)
    resized = img.resize((int(width*scale), int(height*scale)), Image.NEAREST)
    resized = resized.crop((0,0,new_width, new_height))

    return resized


def inference(input_img, transparancy = 0.5, target_layer_number = -1):
    input_img = resize_image_pil(input_img,32,32)
    input_img = np.array(input_img)
    org_img = input_img
    input_img= input_img.reshape((32,32,3))
    transform = transforms.ToTensor()
    input_img = transform(input_img)
    input_img = input_img.unsqueeze(0)
    outputs = model(input_img)
    print(outputs)
    softmax =  torch.nn.Softmax(dim=0)
    o = softmax(outputs.flatten())
    confidences = {classes[i]:float(o[i]) for i in range(10)}
    prediction= torch.max(outputs, 1)
    target_layers = [model.layer2[target_layer_number]]
    cam = GradCAM(model=model, target_layers=target_layers)
    grayscale_cam = cam(input_tensor= input_img)
    grayscale_cam = grayscale_cam[0, :]
    visualization = show_cam_on_image(org_img/255,grayscale_cam,use_rgb=True,
                                     image_weight=transparancy)
    
    return classes[prediction[0].item(),visualization,confidences]
    
demo = gr.Interface(
    inference,
    inputs = [
        gr.Image(width=256,height=256,label="input image"),
        gr.Slider(0,1,value=0.5,label="Overall opacity of the overelay"),
        gr.Slider(-2,-1, value =-2, step=1, label= "Which layer for Gradcam")
    ],
    outputs = [
        "text",
        gr.Image(width= 256, height=256,label="Output"),
        gr.Label(num_top_classes=3)        
    ],
    title = "CIFAR 10 trained on ResNet model in pytorch lightning with Gradcam",
    description = " A simple gradio inference to infer on resnet18 model",
    examples = [["cat.jpg", 0.5, -1],["dog.jpg",0.7,-2]]
)

if __name__ == "__main__":
    demo.launch()