app.py

# common
import os, sys
import math
#import numpy as np

#from random import randrange

# torch
import torch
from torch import nn
#from torch import einsum

import torch.nn.functional as F

#from torch import optim
#from torch.optim import lr_scheduler

#from torch.utils.data import DataLoader
#from torch.utils.data.sampler import SubsetRandomSampler

# torchVision
import torchvision
from torchvision import transforms
#from torchvision import  models
#from torchvision.datasets import CIFAR10, CIFAR100

# torchinfo
#from torchinfo import summary

# Define model
class WideBasic(nn.Module):
    def __init__(self, in_channels, out_channels, stride=1):
        super().__init__()
        self.residual = nn.Sequential(
            nn.BatchNorm2d(in_channels),
            nn.ReLU(inplace=True),
            nn.Conv2d(
                in_channels,
                out_channels,
                kernel_size=3,
                stride=stride,
                padding=1
            ),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(inplace=True),
            nn.Dropout(),
            nn.Conv2d(
                out_channels,
                out_channels,
                kernel_size=3,
                stride=1,
                padding=1
            )
        )

        self.shortcut = nn.Sequential()

        if in_channels != out_channels or stride != 1:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_channels, out_channels, 1, stride=stride)
            )

    def forward(self, x):
        residual = self.residual(x)
        shortcut = self.shortcut(x)

        return residual + shortcut

class WideResNet(nn.Module):
    def __init__(self, num_classes, block, depth=50, widen_factor=1):
        super().__init__()

        self.depth = depth
        k = widen_factor
        l = int((depth - 4) / 6)
        self.in_channels = 16
        self.init_conv = nn.Conv2d(3, self.in_channels, 3, 1, padding=1)
        self.conv2 = self._make_layer(block, 16 * k, l, 1)
        self.conv3 = self._make_layer(block, 32 * k, l, 2)
        self.conv4 = self._make_layer(block, 64 * k, l, 2)
        self.bn = nn.BatchNorm2d(64 * k)
        self.relu = nn.ReLU(inplace=True)
        self.avg_pool = nn.AdaptiveAvgPool2d((1, 1))
        self.linear = nn.Linear(64 * k, num_classes)

    def forward(self, x):
        x = self.init_conv(x)
        x = self.conv2(x)
        x = self.conv3(x)
        x = self.conv4(x)
        x = self.bn(x)
        x = self.relu(x)
        x = self.avg_pool(x)
        x = x.view(x.size(0), -1)
        x = self.linear(x)

        return x

    def _make_layer(self, block, out_channels, num_blocks, stride):
        strides = [stride] + [1] * (num_blocks - 1)
        layers = []
        for stride in strides:
            layers.append(block(self.in_channels, out_channels, stride))
            self.in_channels = out_channels

        return nn.Sequential(*layers)


model = WideResNet(10, WideBasic, depth=40, widen_factor=10)
model.load_state_dict(
        torch.load("weights/cifar10_wide_resnet_model.pt",
            map_location=torch.device('cpu'))
    )

model.eval()

import gradio as gr
from torchvision import transforms

import os
import glob

examples_dir = './examples'
example_files = glob.glob(os.path.join(examples_dir, '*.png'))

normalize = transforms.Normalize(
                            mean=[0.4914, 0.4822, 0.4465],
                            std=[0.2470, 0.2435, 0.2616],
                            )

transform = transforms.Compose([
                                    transforms.ToTensor(),
                                    normalize,
])

classes = [
    "airplane",
    "automobile",
    "bird",
    "cat",
    "deer",
    "dog",
    "frog",
    "horse",
    "ship",
    "truck",
]

def predict(image):
    tsr_image = transform(image).unsqueeze(dim=0)

    model.eval()
    with torch.no_grad():
        pred = model(tsr_image)
        prob = torch.nn.functional.softmax(pred[0], dim=0)

    confidences = {classes[i]: float(prob[i]) for i in range(10)} 
    return confidences


with gr.Blocks(css=".gradio-container {background:honeydew;}", title="WideResNet - CIFAR10 Classification"
               ) as demo:
    gr.HTML("""<div style="font-family:'Times New Roman', 'Serif'; font-size:16pt; font-weight:bold; text-align:center; color:royalblue;">WideResNet - CIFAR10 Classification</div>""")

    with gr.Row():
        input_image = gr.Image(type="pil", image_mode="RGB", shape=(32, 32))

        output_label=gr.Label(label="Probabilities", num_top_classes=3)

    send_btn = gr.Button("Infer")
    
    with gr.Row():
        gr.Examples(['./examples/cifar10_test00.png'], label='Sample images : dog', inputs=input_image)
        gr.Examples(['./examples/cifar10_test01.png'], label='ship', inputs=input_image)
        gr.Examples(['./examples/cifar10_test02.png'], label='airplane', inputs=input_image)
        gr.Examples(['./examples/cifar10_test03.png'], label='frog', inputs=input_image)
        gr.Examples(['./examples/cifar10_test04.png'], label='truck', inputs=input_image)
        gr.Examples(['./examples/cifar10_test05.png'], label='automobile', inputs=input_image)

        #gr.Examples(example_files, inputs=input_image)
        #gr.Examples(['examples/sample02.png', 'examples/sample04.png'], inputs=input_image2)

    send_btn.click(fn=predict, inputs=input_image, outputs=output_label)

# demo.queue(concurrency_count=3)
demo.launch()


### EOF ###