Create app.py

app.py

@@ -0,0 +1,162 @@
+# --------------------------------------------------------
+# YOLOv12 Streamlit App with Emoji-Powered UI 😎🚀
+# Based on yolov10: https://github.com/THU-MIG/yolov10/app.py
+# --------------------------------------------------------
+
+import streamlit as st
+import cv2
+import tempfile
+from ultralytics import YOLO
+from PIL import Image
+import os
+
+# Page config with a cool vibe
+st.set_page_config(
+    page_title="YOLOv12 Detector 🕵️‍♂️",
+    page_icon="🔍",
+    layout="wide"
+)
+
+def yolov12_inference(uploaded_file, model_id, image_size, conf_threshold, input_type):
+    """The magic happens here ✨"""
+    model = YOLO(model_id)
+    
+    if input_type == "Image" and uploaded_file:
+        with st.spinner("🖼️ Painting detections..."):
+            image = Image.open(uploaded_file)
+            results = model.predict(source=image, imgsz=image_size, conf=conf_threshold)
+            annotated_image = results[0].plot()
+            return annotated_image[:, :, ::-1], None
+            
+    elif input_type == "Video" and uploaded_file:
+        with st.spinner("🎥 Cooking up a detected video..."):
+            video_path = tempfile.mktemp(suffix=".mp4")
+            with open(video_path, "wb") as f:
+                f.write(uploaded_file.read())
+
+            cap = cv2.VideoCapture(video_path)
+            fps = cap.get(cv2.CAP_PROP_FPS)
+            frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+            frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+
+            output_video_path = tempfile.mktemp(suffix=".mp4")
+            out = cv2.VideoWriter(output_video_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (frame_width, frame_height))
+
+            frame_count = 0
+            while cap.isOpened():
+                ret, frame = cap.read()
+                if not ret:
+                    break
+                results = model.predict(source=frame, imgsz=image_size, conf=conf_threshold)
+                annotated_frame = results[0].plot()
+                out.write(annotated_frame)
+                frame_count += 1
+                if frame_count % 30 == 0:
+                    st.text(f"🍳 Processed {frame_count} frames...")
+
+            cap.release()
+            out.release()
+            
+            if os.path.exists(video_path):
+                os.remove(video_path)
+                
+            return None, output_video_path
+            
+    return None, None
+
+def main():
+    # Header with flair
+    st.title("YOLOv12: Object Detection Superhero 🦸‍♂️")
+    st.markdown("Powered by xAI's cosmic tech 🌌 | [arXiv 📜](https://arxiv.org/abs/2502.12524) | [GitHub 🐙](https://github.com/sunsmarterjie/yolov12)")
+    
+    # Layout in two columns
+    col1, col2 = st.columns([1, 1])
+    
+    with col1:
+        st.subheader("🎮 Control Room")
+        
+        # Upload section
+        uploaded_file = st.file_uploader(
+            "Drop your file here 📥 - Images or Videos welcome!",
+            type=['jpg', 'jpeg', 'png', 'mp4'],
+            help="Upload an image or video to detect objects in!"
+        )
+        
+        # Input type selector
+        input_type = st.radio(
+            "What's your flavor? 🍦",
+            ("Image", "Video"),
+            help="Tell me if it's a still or moving picture!"
+        )
+        
+        # Model selection
+        model_id = st.selectbox(
+            "Pick your YOLO weapon ⚔️",
+            ["yolov12n.pt", "yolov12s.pt", "yolov12m.pt", "yolov12l.pt", "yolov12x.pt"],
+            index=2,
+            help="Choose your model power level: n (nano) to x (extra spicy)!"
+        )
+        
+        # Image size slider
+        image_size = st.slider(
+            "Zoom level 🔍",
+            min_value=320,
+            max_value=1280,
+            value=640,
+            step=32,
+            help="Bigger numbers = sharper eyes (but slower)!"
+        )
+        
+        # Confidence threshold
+        conf_threshold = st.slider(
+            "Certainty meter 🎯",
+            min_value=0.0,
+            max_value=1.0,
+            value=0.25,
+            step=0.05,
+            help="How sure should I be? Higher = pickier!"
+        )
+        
+        # The big red button
+        if st.button("Detect Objects! 🚀", help="Click to unleash the detection magic!"):
+            if uploaded_file is None:
+                st.error("Yo! Upload something first 🙈")
+            else:
+                annotated_image, annotated_video = yolov12_inference(
+                    uploaded_file, model_id, image_size, conf_threshold, input_type
+                )
+                st.session_state['results'] = (annotated_image, annotated_video)
+    
+    with col2:
+        st.subheader("🖥️ Detection HQ")
+        
+        # Display results
+        if 'results' in st.session_state:
+            annotated_image, annotated_video = st.session_state['results']
+            
+            if input_type == "Image" and annotated_image is not None:
+                st.image(
+                    annotated_image,
+                    caption="Your Detected Masterpiece 🎨",
+                    use_column_width=True
+                )
+            elif input_type == "Video" and annotated_video is not None:
+                st.video(
+                    annotated_video,
+                    format="video/mp4",
+                    start_time=0
+                )
+                # Clean up temporary video file
+                if os.path.exists(annotated_video):
+                    os.remove(annotated_video)
+            else:
+                st.warning("Nothing to show yet! Hit the button! ⚡")
+        else:
+            st.info("Awaiting your command, captain! 🖖 Upload and detect to see results!")
+    
+    # Footer with sass
+    st.markdown("---")
+    st.markdown("Built with 💖 by xAI's minions | Objects beware, YOLOv12 is here! 😈")
+
+if __name__ == '__main__':
+    main()