Update app.py

app.py

@@ -1,52 +1,17 @@
-import torch
 import gradio as gr
 import numpy as np
-from PIL import Image
-import torchvision.transforms as T
+import random
 
-# 假設模型是分類用簡單 CNN（你要換成你的模型定義）
-class SimpleCNN(torch.nn.Module):
-    def __init__(self):
-        super().__init__()
-        self.conv = torch.nn.Conv2d(3, 16, 3, stride=2)
-        self.fc = torch.nn.Linear(16*111*111, 2)  # 2分類為例
-
-    def forward(self, x):
-        x = self.conv(x)
-        x = torch.relu(x)
-        x = x.view(x.size(0), -1)
-        x = self.fc(x)
-        return torch.softmax(x, dim=1)
-
-# 載入模型與權重
-model = SimpleCNN()
-model.load_state_dict(torch.load("salg_model.pt", map_location="cpu"))
-model.eval()
-
-# 影像預處理
-transform = T.Compose([
-    T.Resize((224, 224)),
-    T.ToTensor(),
-])
-
-def predict(image):
-    img = Image.fromarray(image.astype('uint8'), 'RGB')
-    input_tensor = transform(img).unsqueeze(0)  # 加 batch 維度
-    with torch.no_grad():
-        pred = model(input_tensor)
-    # 取最高機率標籤
-    pred_label = torch.argmax(pred, dim=1).item()
-    confidence = pred[0][pred_label].item()
-    # 在圖上標示結果（簡單做法）
-    result_img = image.copy()
-    import cv2
-    cv2.putText(result_img, f"Label: {pred_label} Conf: {confidence:.2f}",
-                (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (0,255,0), 2)
-    return result_img, f"Label: {pred_label}, Confidence: {confidence:.2f}"
+# 模擬辨識按鈕的回傳邏輯
+def fake_detect(image):
+    # 隨機回傳 "yap" 或 "not"
+    result = random.choice(["yap", "not"])
+    return image, result
 
+# 建立 Gradio 介面
 with gr.Blocks() as demo:
-    gr.Markdown("## 🦺 Helmet Detector with Model")
-    gr.Markdown("請上傳圖片，點擊「辨識」由模型進行預測")
+    gr.Markdown("## 🦺 Helmet ")
+    gr.Markdown("請上傳一張圖片，然後點擊「辨識」按鈕模擬結果展示")
 
     with gr.Row():
         with gr.Column():
@@ -55,8 +20,8 @@ with gr.Blocks() as demo:
 
         with gr.Column():
             image_output = gr.Image(type="numpy", label="推論結果")
-            result_label = gr.Textbox(label="辨識摘要")
+            result_label = gr.Textbox(label="辨識摘要", placeholder="結果")
 
-    detect_button.click(fn=predict, inputs=image_input, outputs=[image_output, result_label])
+    detect_button.click(fn=fake_detect, inputs=image_input, outputs=[image_output, result_label])
 
 demo.launch()