Create app.py

app.py

@@ -0,0 +1,128 @@
+# ── Gradio 설치 ──────────────────────────
+!pip install gradio -q
+
+import gradio as gr
+import torch
+import torch.nn as nn
+import torchvision.transforms as transforms
+from PIL import Image
+from datasets import load_dataset
+
+# ── 클래스 이름 가져오기 ──────────────────
+dataset = load_dataset("food101", split="train[:1%]")
+class_names = dataset.features['label'].names
+
+# ── 모델 정의 (똑같이 붙여넣기) ───────────
+class BottleneckBlock(nn.Module):
+    expansion = 4
+
+    def __init__(self, in_channels, mid_channels, stride=1):
+        super().__init__()
+        self.conv1 = nn.Conv2d(in_channels, mid_channels, kernel_size=1, bias=False)
+        self.bn1   = nn.BatchNorm2d(mid_channels)
+        self.conv2 = nn.Conv2d(mid_channels, mid_channels, kernel_size=3,
+                               stride=stride, padding=1, bias=False)
+        self.bn2   = nn.BatchNorm2d(mid_channels)
+        self.conv3 = nn.Conv2d(mid_channels, mid_channels * self.expansion,
+                               kernel_size=1, bias=False)
+        self.bn3   = nn.BatchNorm2d(mid_channels * self.expansion)
+        self.relu  = nn.ReLU(inplace=True)
+        self.shortcut = nn.Sequential()
+        if stride != 1 or in_channels != mid_channels * self.expansion:
+            self.shortcut = nn.Sequential(
+                nn.Conv2d(in_channels, mid_channels * self.expansion,
+                          kernel_size=1, stride=stride, bias=False),
+                nn.BatchNorm2d(mid_channels * self.expansion)
+            )
+
+    def forward(self, x):
+        identity = x
+        out = self.relu(self.bn1(self.conv1(x)))
+        out = self.relu(self.bn2(self.conv2(out)))
+        out = self.bn3(self.conv3(out))
+        out += self.shortcut(identity)
+        out = self.relu(out)
+        return out
+
+class ResNet50(nn.Module):
+    def __init__(self, num_classes=101):
+        super().__init__()
+        self.conv1   = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3, bias=False)
+        self.bn1     = nn.BatchNorm2d(64)
+        self.relu    = nn.ReLU(inplace=True)
+        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
+        self.layer1  = self._make_layer( 64,  64, blocks=3, stride=1)
+        self.layer2  = self._make_layer(256, 128, blocks=4, stride=2)
+        self.layer3  = self._make_layer(512, 256, blocks=6, stride=2)
+        self.layer4  = self._make_layer(1024, 512, blocks=3, stride=2)
+        self.avgpool = nn.AdaptiveAvgPool2d((1, 1))
+        self.fc      = nn.Linear(2048, num_classes)
+
+    def _make_layer(self, in_channels, mid_channels, blocks, stride):
+        layers = [BottleneckBlock(in_channels, mid_channels, stride=stride)]
+        for _ in range(1, blocks):
+            layers.append(BottleneckBlock(mid_channels * 4, mid_channels))
+        return nn.Sequential(*layers)
+
+    def forward(self, x):
+        x = self.maxpool(self.relu(self.bn1(self.conv1(x))))
+        x = self.layer1(x)
+        x = self.layer2(x)
+        x = self.layer3(x)
+        x = self.layer4(x)
+        x = self.avgpool(x)
+        x = torch.flatten(x, 1)
+        x = self.fc(x)
+        return x
+
+# ── 모델 불러오기 ─────────────────────────
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+model = ResNet50(num_classes=101)
+model.load_state_dict(
+    torch.load('resnet50_food101.pth', map_location=device)
+)
+model = model.to(device)
+model.eval()
+
+# ── 전처리 ────────────────────────────────
+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]
+    )
+])
+
+# ── 예측 함수 ─────────────────────────────
+def predict(image):
+    # PIL 이미지 → 텐서 변환
+    img_tensor = transform(image).unsqueeze(0).to(device)
+    # unsqueeze(0) = (3,224,224) → (1,3,224,224) 배치 차원 추가
+
+    with torch.no_grad():
+        output = model(img_tensor)
+        probs  = torch.softmax(output, dim=1)  # 점수 → 확률
+        top5   = probs.topk(5)                 # 상위 5개
+
+    # 결과 딕셔너리로 반환
+    result = {}
+    for i in range(5):
+        label = class_names[top5.indices[0][i].item()]
+        prob  = top5.values[0][i].item()
+        result[label] = prob
+
+    return result
+
+# ── Gradio 인터페이스 ──────────────────────
+demo = gr.Interface(
+    fn=predict,                          # 예측 함수
+    inputs=gr.Image(type="pil"),         # 입력: 이미지
+    outputs=gr.Label(num_top_classes=5), # 출력: 상위 5개 클래스
+    title="🍔 Food-101 분류기",
+    description="음식 사진��� 올리면 어떤 음식인지 맞춰줘요! (101가지 음식 분류)",
+    examples=[],                         # 예시 이미지 (있으면 추가)
+)
+
+demo.launch()