Update app.py

app.py

@@ -9,11 +9,12 @@ from gtts import gTTS
 # ===============================
 processor = None
 caption_model = None
+yolo_model = None
 
 device = "cpu"
 
 # ===============================
-# LOAD MODEL
+# LOAD BLIP
 # ===============================
 def load_blip():
     global processor, caption_model
@@ -29,7 +30,16 @@ def load_blip():
         ).to(device)
 
 # ===============================
-# GENERATE CAPTION
+# LOAD YOLO
+# ===============================
+def load_yolo():
+    global yolo_model
+    if yolo_model is None:
+        from ultralytics import YOLO
+        yolo_model = YOLO("yolov8n.pt")  # lightweight model
+
+# ===============================
+# CAPTION
 # ===============================
 def generate_caption(image):
     load_blip()
@@ -41,7 +51,25 @@ def generate_caption(image):
     return caption
 
 # ===============================
-# TRANSLATE + AUDIO (SAFE)
+# OBJECT DETECTION
+# ===============================
+def detect_objects(image):
+    load_yolo()
+
+    results = yolo_model(image)
+
+    labels = []
+    for r in results:
+        for box in r.boxes:
+            cls = int(box.cls[0])
+            name = yolo_model.names[cls]
+            conf = float(box.conf[0])
+            labels.append(f"{name} ({conf:.2f})")
+
+    return ", ".join(labels) if labels else "No objects detected"
+
+# ===============================
+# TRANSLATE + AUDIO
 # ===============================
 def translate_and_speak(text, selected_languages):
 
@@ -65,7 +93,6 @@ def translate_and_speak(text, selected_languages):
         if not code:
             continue
 
-        # 🔤 Translation safe
         try:
             translated = GoogleTranslator(source='auto', target=code).translate(text)
         except:
@@ -76,7 +103,6 @@ def translate_and_speak(text, selected_languages):
 
         translations_text += f"{lang}: {translated}\n\n"
 
-        # 🔊 Text-to-Speech safe
         try:
             file_path = f"{lang}.mp3"
             tts = gTTS(translated, lang=code)
@@ -93,40 +119,46 @@ def translate_and_speak(text, selected_languages):
 def process_image(image, selected_languages):
 
     if image is None:
-        return "⚠️ Please upload an image.", "", None
+        return "⚠️ Upload image", "", "", None
 
     caption = generate_caption(image)
+    detections = detect_objects(image)
 
     translations, audios = translate_and_speak(caption, selected_languages)
 
-    return caption, translations, audios if audios else None
+    return caption, detections, translations, audios if audios else None
 
 # ===============================
-# UI
+# UI (PREMIUM)
 # ===============================
-with gr.Blocks() as demo:
+with gr.Blocks(theme=gr.themes.Soft(), css="""
+#title {text-align: center; font-size: 30px; font-weight: bold;}
+.card {padding: 15px; border-radius: 15px; box-shadow: 0px 2px 10px rgba(0,0,0,0.1);}
+""") as demo:
 
-    gr.Markdown("## 🚀 AI Image Recognition (Multi-Language + Voice)")
+    gr.Markdown("<div id='title'>🚀 AI Vision Assistant</div>")
 
-    input_image = gr.Image(type="pil", label="Upload Image")
-
-    language_select = gr.CheckboxGroup(
-        choices=["Telugu", "Hindi", "Tamil", "Kannada", "Malayalam"],
-        label="Select Languages"
-    )
+    with gr.Row():
+        with gr.Column(scale=1):
+            input_image = gr.Image(type="pil", label="📸 Upload Image")
 
-    btn = gr.Button("Analyze Image")
+            language_select = gr.CheckboxGroup(
+                ["Telugu", "Hindi", "Tamil", "Kannada", "Malayalam"],
+                label="🌍 Select Languages"
+            )
 
-    caption_out = gr.Textbox(label="English Caption")
-    translations_out = gr.Textbox(label="Translations")
+            btn = gr.Button("🔍 Analyze", variant="primary")
 
-    # safer than Audio for multiple files
-    audio_out = gr.File(label="Download Audio Files")
+        with gr.Column(scale=1):
+            caption_out = gr.Textbox(label="🧠 Caption", elem_classes="card")
+            detection_out = gr.Textbox(label="🎯 Detected Objects", elem_classes="card")
+            translations_out = gr.Textbox(label="🌍 Translations", elem_classes="card")
+            audio_out = gr.File(label="🔊 Audio Files")
 
     btn.click(
         process_image,
         inputs=[input_image, language_select],
-        outputs=[caption_out, translations_out, audio_out]
+        outputs=[caption_out, detection_out, translations_out, audio_out]
     )
 
 # ===============================