Add files via upload

synais8hmata/app .py

@@ -0,0 +1,321 @@
+# =========================================================
+# ΜΕΡΟΣ Β: FLASK WEB APP ΓΙΑ ΤΗΝ ΕΡΓΑΣΙΑ
+# Αποθήκευσέ το ως app.py
+# =========================================================
+
+import os
+import cv2
+from flask import Flask, request, render_template_string, send_from_directory
+from fer import FER
+from openai import OpenAI
+
+# =========================================================
+# ΡΥΘΜΙΣΕΙΣ OPENAI (ΠΑΛΙΟ ΙΔΙΟ CONCEPT)
+# =========================================================
+
+OPENAI_API_KEY = os.environ.get("OPENAI_API_KEY", "")  # ή βάλε σταθερά αν θες
+client = None
+if OPENAI_API_KEY:
+    client = OpenAI(api_key=OPENAI_API_KEY)
+
+LLM_MODEL = "gpt-4.1-mini"
+
+# =========================================================
+# ΔΗΜΙΟΥΡΓΙΑ FLASK APP
+# =========================================================
+
+app = Flask(__name__)
+
+# Φάκελοι για uploads / annotated
+UPLOAD_FOLDER = "static/uploads"
+ANNOTATED_FOLDER = "static/annotated"
+
+os.makedirs(UPLOAD_FOLDER, exist_ok=True)
+os.makedirs(ANNOTATED_FOLDER, exist_ok=True)
+
+# Δημιουργία detector
+emotion_detector = FER(mtcnn=True)
+
+# =========================================================
+# ΒΟΗΘΗΤΙΚΕΣ ΣΥΝΑΡΤΗΣΕΙΣ (ΙΔΙΕΣ ΛΟΓΙΚΑ ΜΕ ΤΟ NOTEBOOK)
+# =========================================================
+
+def allowed_emotions():
+    return ["angry", "disgust", "fear", "happy", "sad", "surprise", "neutral"]
+
+
+def detect_emotions_in_image(image_path: str):
+    img_bgr = cv2.imread(image_path)
+    if img_bgr is None:
+        raise FileNotFoundError(f"Η εικόνα δεν βρέθηκε: {image_path}")
+
+    img_rgb = cv2.cvtColor(img_bgr, cv2.COLOR_BGR2RGB)
+    raw = emotion_detector.detect_emotions(img_rgb)
+
+    results = []
+    for det in raw:
+        box = det.get("box", [0, 0, 0, 0])
+        emo = det.get("emotions", {})
+        if emo:
+            label = max(emo, key=emo.get)
+            conf = float(emo[label])
+        else:
+            label = "unknown"
+            conf = 0.0
+
+        results.append({
+            "box": tuple(box),
+            "emotion": label,
+            "confidence": conf,
+            "emotions": {k: float(v) for k, v in emo.items()}
+        })
+    return results
+
+
+def aggregate_emotions(detections):
+    base = {e: 0.0 for e in allowed_emotions()}
+    if not detections:
+        return base
+
+    count = 0
+    for det in detections:
+        emo = det["emotions"]
+        if emo:
+            count += 1
+            for e in allowed_emotions():
+                base[e] += float(emo.get(e, 0.0))
+
+    if count == 0:
+        return base
+
+    for e in base:
+        base[e] /= count
+
+    return base
+
+
+def annotate_image(image_path: str, detections, save_path: str):
+    """
+    Φορτώνει την εικόνα, ζωγραφίζει τα κουτιά και την αποθηκεύει στο save_path.
+    """
+    img = cv2.imread(image_path)
+    if img is None:
+        raise FileNotFoundError("Αποτυχία φόρτωσης εικόνας.")
+
+    for det in detections:
+        x, y, w, h = det["box"]
+        emotion = det["emotion"]
+        conf = det["confidence"]
+
+        cv2.rectangle(img, (x, y), (x + w, y + h), (0, 255, 0), 2)
+        label = f"{emotion} ({conf:.2f})"
+        cv2.putText(
+            img,
+            label,
+            (x, y - 10),
+            cv2.FONT_HERSHEY_SIMPLEX,
+            0.6,
+            (0, 255, 0),
+            2,
+            cv2.LINE_AA
+        )
+
+    cv2.imwrite(save_path, img)
+
+
+emotion_to_playlist = {
+    "happy": "uplifting_pop_playlist",
+    "sad": "chill_ambient_playlist",
+    "angry": "rock_energy_playlist",
+    "fear": "calm_piano_playlist",
+    "disgust": "neutral_focus_playlist",
+    "surprise": "random_discovery_playlist",
+    "neutral": "lofi_study_playlist",
+}
+
+def map_emotion_to_playlist(dominant_emotion: str) -> str:
+    return emotion_to_playlist.get(dominant_emotion, "lofi_study_playlist")
+
+
+def llm_description(summary: dict):
+    if client is None:
+        return "LLM δεν κλήθηκε (δεν έχει οριστεί OPENAI_API_KEY)."
+
+    text = ", ".join([f"{e}: {v:.2f}" for e, v in summary.items()])
+
+    prompt = (
+        "You are an assistant that describes emotional state based on probabilities.\n"
+        f"Emotion probabilities: {text}\n"
+        "Write a short, clear emotional interpretation in English."
+    )
+
+    try:
+        response = client.chat.completions.create(
+            model=LLM_MODEL,
+            messages=[
+                {"role": "system", "content": "You explain emotional states accurately."},
+                {"role": "user", "content": prompt}
+            ],
+            max_tokens=120,
+            temperature=0.4
+        )
+        return response.choices[0].message.content.strip()
+    except Exception as e:
+        return f"LLM error: {e}"
+
+
+# =========================================================
+# ΑΠΛΟ HTML TEMPLATE ΣΕ CHAT-STYLE
+# =========================================================
+
+HTML_TEMPLATE = """
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <title>Emotion-Based Music Assistant</title>
+    <style>
+        body { font-family: Arial, sans-serif; max-width: 900px; margin: auto; padding: 20px; }
+        .chat-box { border: 1px solid #ccc; border-radius: 8px; padding: 15px; margin-top: 20px; }
+        .message-user { text-align: right; color: #333; margin: 10px 0; }
+        .message-bot { text-align: left; color: #006; margin: 10px 0; }
+        .image-row { display: flex; gap: 20px; margin-top: 20px; }
+        .image-row img { max-width: 100%; max-height: 300px; border-radius: 8px; border: 1px solid #ccc; }
+        .result-block { margin-top: 20px; padding: 10px; border-radius: 8px; background-color: #f8f8f8; }
+    </style>
+</head>
+<body>
+    <h1>Emotion-Based Music Assistant</h1>
+    <p>Ανέβασε μία εικόνα προσώπου. Το σύστημα θα ανιχνεύσει το συναίσθημα, θα προτείνει playlist και θα δώσει περιγραφή μέσω LLM.</p>
+
+    <form action="/analyze" method="post" enctype="multipart/form-data">
+        <label>Εικόνα:</label>
+        <input type="file" name="image" accept="image/*" required>
+        <button type="submit">Ανάλυση</button>
+    </form>
+
+    {% if user_message %}
+    <div class="chat-box">
+        <div class="message-user">
+            <strong>Εσύ:</strong> {{ user_message }}
+        </div>
+        <div class="message-bot">
+            <strong>Σύστημα (Dominant emotion):</strong> {{ dominant_emotion }}
+        </div>
+        <div class="message-bot">
+            <strong>Προτεινόμενη playlist:</strong> {{ playlist }}
+        </div>
+        <div class="message-bot">
+            <strong>LLM περιγραφή:</strong> {{ llm_text }}
+        </div>
+
+        <div class="result-block">
+            <h3>Μέσες πιθανότητες συναισθημάτων:</h3>
+            <ul>
+                {% for emo, prob in summary.items() %}
+                    <li>{{ emo }}: {{ "%.2f"|format(prob) }}</li>
+                {% endfor %}
+            </ul>
+        </div>
+
+        <div class="image-row">
+            <div>
+                <h3>Αρχική εικόνα</h3>
+                <img src="{{ original_url }}" alt="Original Image">
+            </div>
+            <div>
+                <h3>Annotated εικόνα</h3>
+                <img src="{{ annotated_url }}" alt="Annotated Image">
+            </div>
+        </div>
+    </div>
+    {% endif %}
+</body>
+</html>
+"""
+
+
+# =========================================================
+# ROUTES
+# =========================================================
+
+@app.route("/", methods=["GET"])
+def index():
+    """
+    Αρχική σελίδα με φόρμα upload.
+    """
+    return render_template_string(
+        HTML_TEMPLATE,
+        user_message=None,
+        dominant_emotion=None,
+        playlist=None,
+        llm_text=None,
+        summary={},
+        original_url="",
+        annotated_url=""
+    )
+
+
+@app.route("/analyze", methods=["POST"])
+def analyze():
+    """
+    Route που δέχεται την εικόνα, τρέχει FER + LLM και επιστρέφει αποτελέσματα.
+    """
+    file = request.files.get("image")
+    if not file:
+        return "Δεν δόθηκε εικόνα.", 400
+
+    # Αποθήκευση αρχικής εικόνας
+    filename = file.filename
+    original_path = os.path.join(UPLOAD_FOLDER, filename)
+    file.save(original_path)
+
+    # Ανάλυση συναισθημάτων
+    detections = detect_emotions_in_image(original_path)
+    summary = aggregate_emotions(detections)
+
+    if detections:
+        dominant = max(summary, key=summary.get)
+    else:
+        dominant = "none"
+
+    playlist = map_emotion_to_playlist(dominant)
+    llm_text = llm_description(summary)
+
+    # Δημιουργία annotated εικόνας
+    annotated_filename = f"{os.path.splitext(filename)[0]}_annotated.png"
+    annotated_path = os.path.join(ANNOTATED_FOLDER, annotated_filename)
+
+    if detections:
+        annotate_image(original_path, detections, annotated_path)
+    else:
+        # αν δεν υπάρχουν πρόσωπα, αντιγράφουμε απλά την αρχική
+        cv2.imwrite(annotated_path, cv2.imread(original_path))
+
+    # URLs για εμφάνιση εικόνων
+    original_url = f"/static/uploads/{filename}"
+    annotated_url = f"/static/annotated/{annotated_filename}"
+
+    # Μήνυμα χρήστη (σαν chat)
+    user_message = f"Ανέβασα την εικόνα {filename}."
+
+    return render_template_string(
+        HTML_TEMPLATE,
+        user_message=user_message,
+        dominant_emotion=dominant,
+        playlist=playlist,
+        llm_text=llm_text,
+        summary=summary,
+        original_url=original_url,
+        annotated_url=annotated_url
+    )
+
+
+# =========================================================
+# MAIN
+# =========================================================
+
+if __name__ == "__main__":
+    # Σημείωση: Για production deployment σε Render/PythonAnywhere,
+    # συνήθως ΔΕΝ τρέχουμε debug=True
+    app.run(host="0.0.0.0", port=8000, debug=True)

synais8hmata/requirements.txt

@@ -0,0 +1,6 @@
+flask
+fer==22.4.0
+opencv-python-headless
+openai
+numpy
+pillow