Create app.py

app.py

@@ -0,0 +1,94 @@
+import cv2
+import time
+import csv
+import gradio as gr
+import numpy as np
+import sounddevice as sd
+import soundfile as sf
+from deepface import DeepFace
+import tempfile
+import os
+from scipy.io.wavfile import write as write_wav
+
+
+# === Audio recording config ===
+SAMPLE_RATE = 16000
+DURATION = 30  # seconds
+
+
+def record_audio(filename="audio_recording.wav"):
+    print("Recording audio for 30 seconds...")
+    audio = sd.rec(int(SAMPLE_RATE * DURATION), samplerate=SAMPLE_RATE, channels=1, dtype='int16')
+    sd.wait()
+    write_wav(filename, SAMPLE_RATE, audio)
+    return filename
+
+
+def map_emotions_to_metrics(emotion):
+    valence = emotion.get("happy", 0) / 100
+    arousal = (emotion.get("angry", 0) + emotion.get("surprise", 0)) / 200
+    stress_index = emotion.get("fear", 0) / 100
+    engagement_level = 1 - emotion.get("neutral", 0) / 100
+    dominance = (emotion.get("angry", 0) + emotion.get("disgust", 0)) / 200
+    return valence, arousal, dominance, stress_index, engagement_level
+
+
+def analyze_inputs():
+    cap = cv2.VideoCapture(0)
+    start_time = time.time()
+    data = []
+
+    # Start audio recording in parallel
+    audio_file_path = os.path.join(tempfile.gettempdir(), "temp_audio.wav")
+    record_audio(audio_file_path)
+
+    for i in range(DURATION):
+        ret, frame = cap.read()
+        if not ret:
+            continue
+
+        try:
+            result = DeepFace.analyze(frame, actions=['emotion'], enforce_detection=False)
+            emotion = result[0]['emotion']
+            valence, arousal, dominance, stress_index, engagement_level = map_emotions_to_metrics(emotion)
+
+            data.append([
+                round(time.time() - start_time, 2),
+                round(valence, 3),
+                round(arousal, 3),
+                round(dominance, 3),
+                round(stress_index, 3),
+                round(engagement_level, 3)
+            ])
+
+        except Exception as e:
+            print("Error analyzing frame:", e)
+        time.sleep(1)
+
+    cap.release()
+
+    # Write to CSV
+    csv_path = os.path.join(tempfile.gettempdir(), "emotional_metrics.csv")
+    with open(csv_path, mode='w', newline='') as f:
+        writer = csv.writer(f)
+        writer.writerow(["timestamp", "valence", "arousal", "dominance", "stress_index", "engagement_level"])
+        writer.writerows(data)
+
+    return csv_path
+
+
+def demo_interface(campaign_text):
+    csv_output = analyze_inputs()
+    return csv_output
+
+
+demo = gr.Interface(
+    fn=demo_interface,
+    inputs=gr.Textbox(label="Enter Campaign Description", lines=2, placeholder="Describe your campaign..."),
+    outputs=gr.File(label="Download Emotional Metrics (CSV Output)"),
+    title="📊 Complex Campaign Emotion Response Engine",
+    description="This demo captures webcam and microphone input for 30 seconds. It analyzes facial expressions using DeepFace and records audio. The output is a downloadable CSV of emotional metrics over time."
+)
+
+if __name__ == "__main__":
+    demo.launch()