Update app.py

app.py

@@ -1,16 +1,15 @@
 import streamlit as st
+from PIL import Image
 import torch
+from torchvision import transforms
 from torchvision.models.detection import fasterrcnn_resnet50_fpn
-from torchvision.transforms import functional as F
-from PIL import Image, ImageDraw, ImageFont
-import io
-import cv2
-import numpy as np
+import torchvision
 
-# Set up UI
-st.set_page_config(page_title="🎯 AI Object Detection", layout="centered")
+# Load model
+model = fasterrcnn_resnet50_fpn(pretrained=True)
+model.eval()
 
-# COCO class labels (80 classes)
+# Define class labels
 COCO_INSTANCE_CATEGORY_NAMES = [
     '__background__', 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus',
     'train', 'truck', 'boat', 'traffic light', 'fire hydrant', 'N/A', 'stop sign',
@@ -20,79 +19,74 @@ COCO_INSTANCE_CATEGORY_NAMES = [
     'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard',
     'surfboard', 'tennis racket', 'bottle', 'N/A', 'wine glass', 'cup', 'fork',
     'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', 'broccoli',
-    'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
-    'potted plant', 'bed', 'N/A', 'dining table', 'N/A', 'N/A', 'toilet', 'N/A',
-    'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave',
-    'oven', 'toaster', 'sink', 'refrigerator', 'N/A', 'book', 'clock', 'vase',
-    'scissors', 'teddy bear', 'hair drier', 'toothbrush'
+    'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch', 'potted plant',
+    'bed', 'N/A', 'dining table', 'N/A', 'N/A', 'toilet', 'N/A', 'tv', 'laptop',
+    'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', 'oven', 'toaster',
+    'sink', 'refrigerator', 'N/A', 'book', 'clock', 'vase', 'scissors', 'teddy bear',
+    'hair drier', 'toothbrush'
 ]
 
-# Load detection model
-@st.cache_resource
-def load_model():
-    model = fasterrcnn_resnet50_fpn(pretrained=True)
-    model.eval()
-    return model
+def get_prediction(img, threshold):
+    transform = transforms.Compose([transforms.ToTensor()])
+    img = transform(img)
+    pred = model([img])
+    pred_classes = [COCO_INSTANCE_CATEGORY_NAMES[i] for i in list(pred[0]['labels'].numpy())]
+    pred_boxes = [[(i[0], i[1]), (i[2], i[3])] for i in list(pred[0]['boxes'].detach().numpy())]
+    pred_score = list(pred[0]['scores'].detach().numpy())
+    pred_t = [pred_score.index(x) for x in pred_score if x > threshold][-1]
+    boxes = pred_boxes[:pred_t+1]
+    classes = pred_classes[:pred_t+1]
+    return boxes, classes
 
-model = load_model()
+# UI design
+st.set_page_config(page_title="AI Object Detector", layout="wide")
 
-# Sidebar
-st.sidebar.title("⚙️ Settings")
-conf_thresh = st.sidebar.slider("Confidence Threshold", 0.2, 1.0, 0.5, 0.05)
-theme = st.sidebar.radio("Theme", ["Light", "Dark"])
-source = st.sidebar.radio("Input Source", ["Upload Image", "Use Webcam"])
+st.markdown("""
+    <style>
+        .main {
+            background-color: #f5f7fa;
+            padding: 20px;
+            border-radius: 10px;
+        }
+        h1 {
+            color: #2c3e50;
+        }
+        .stButton>button {
+            background-color: #008CBA;
+            color: white;
+            font-weight: bold;
+            border-radius: 8px;
+            padding: 10px 24px;
+        }
+    </style>
+""", unsafe_allow_html=True)
 
-# Main Title
-st.title("🎯 Object Detection with AI")
-st.markdown("Upload an image or use webcam to detect objects in real time using **Faster R-CNN** with bounding boxes and labels.")
+st.title("🔍 AI Object Detection App")
+st.markdown("Upload an image and let the AI detect what's in it!")
 
-def detect_objects(image, threshold=0.5):
-    tensor = F.to_tensor(image).unsqueeze(0)
-    outputs = model(tensor)
-    boxes = outputs[0]['boxes']
-    labels = outputs[0]['labels']
-    scores = outputs[0]['scores']
-    
-    draw = ImageDraw.Draw(image)
+img_file = st.file_uploader("📸 Upload an Image", type=["jpg", "jpeg", "png"])
 
-    for box, label, score in zip(boxes, labels, scores):
-        if score >= threshold:
-            box = box.tolist()
-            name = COCO_INSTANCE_CATEGORY_NAMES[label.item()]
-            draw.rectangle(box, outline="red", width=3)
-            draw.text((box[0], box[1] - 10), f"{name} {score:.2f}", fill="white" if theme == "Dark" else "black")
-    return image
+confidence = st.slider("🎯 Confidence Threshold", 0.0, 1.0, 0.5)
 
-if source == "Upload Image":
-    uploaded_file = st.file_uploader("📤 Upload Image", type=["jpg", "jpeg", "png"])
-    if uploaded_file:
-        image = Image.open(uploaded_file).convert("RGB")
-        st.image(image, caption="Original Image", use_column_width=True)
-        st.subheader("🧠 Detected Objects:")
-        detected = detect_objects(image.copy(), conf_thresh)
-        st.image(detected, caption="Detection Result", use_column_width=True)
+if img_file is not None:
+    image = Image.open(img_file).convert("RGB")
+    st.image(image, caption="Uploaded Image", use_column_width=True)
 
-        # Download button
-        buf = io.BytesIO()
-        detected.save(buf, format="PNG")
-        st.download_button("📥 Download Image", data=buf.getvalue(), file_name="detected.png", mime="image/png")
+    boxes, classes = get_prediction(image, confidence)
 
-else:
-    st.subheader("📸 Real-Time Webcam Detection")
-    run = st.checkbox("Turn On Webcam")
+    # Draw results
+    import matplotlib.pyplot as plt
+    import matplotlib.patches as patches
 
-    FRAME_WINDOW = st.image([])
+    fig, ax = plt.subplots(1, figsize=(12, 8))
+    ax.imshow(image)
+    for i in range(len(boxes)):
+        box = boxes[i]
+        label = classes[i]
+        rect = patches.Rectangle(box[0], box[1][0]-box[0][0], box[1][1]-box[0][1],
+                                 linewidth=2, edgecolor='blue', facecolor='none')
+        ax.add_patch(rect)
+        ax.text(box[0][0], box[0][1]-10, label, fontsize=12,
+                color='black', bbox=dict(facecolor='lightblue', edgecolor='blue', boxstyle='round,pad=0.5'))
+    st.pyplot(fig)
 
-    cap = cv2.VideoCapture(0)
-
-    while run:
-        ret, frame = cap.read()
-        if not ret:
-            st.write("Failed to get webcam feed.")
-            break
-
-        img = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
-        pil_img = Image.fromarray(img)
-        result = detect_objects(pil_img.copy(), conf_thresh)
-        FRAME_WINDOW.image(result)
-    cap.release()