Spaces:
Sleeping
Sleeping
updating app
Browse files
app.py
CHANGED
|
@@ -7,6 +7,8 @@ from transformers import pipeline
|
|
| 7 |
from PIL import Image
|
| 8 |
import torch
|
| 9 |
from collections import deque
|
|
|
|
|
|
|
| 10 |
|
| 11 |
# Set page config
|
| 12 |
st.set_page_config(
|
|
@@ -18,8 +20,8 @@ st.set_page_config(
|
|
| 18 |
# --- App Title and Description ---
|
| 19 |
st.title("Advanced Real-Time Emotion Detection")
|
| 20 |
st.write("""
|
| 21 |
-
This app detects emotions in real-time using
|
| 22 |
-
|
| 23 |
""")
|
| 24 |
|
| 25 |
# --- Load Models ---
|
|
@@ -51,10 +53,10 @@ selected_model = st.sidebar.selectbox(
|
|
| 51 |
list(model_options.keys())
|
| 52 |
)
|
| 53 |
|
| 54 |
-
# Input method selection
|
| 55 |
input_method = st.sidebar.radio(
|
| 56 |
"Choose Input Method",
|
| 57 |
-
["Real-time Webcam", "Upload an Image", "Capture Image"]
|
| 58 |
)
|
| 59 |
|
| 60 |
# Confidence threshold
|
|
@@ -69,8 +71,19 @@ confidence_threshold = st.sidebar.slider(
|
|
| 69 |
# Face detection toggle
|
| 70 |
use_face_detection = st.sidebar.checkbox("Enable Face Detection", value=True)
|
| 71 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
# History length for real-time tracking
|
| 73 |
-
if input_method
|
| 74 |
history_length = st.sidebar.slider(
|
| 75 |
"Emotion History Length (seconds)",
|
| 76 |
min_value=5,
|
|
@@ -161,117 +174,79 @@ def draw_faces_with_emotions(image, faces, emotions):
|
|
| 161 |
|
| 162 |
return img
|
| 163 |
|
| 164 |
-
|
| 165 |
-
|
| 166 |
-
|
|
|
|
| 167 |
|
| 168 |
-
|
| 169 |
-
|
| 170 |
-
|
| 171 |
-
|
| 172 |
-
col1, col2 = st.columns(2)
|
| 173 |
-
with col1:
|
| 174 |
-
st.image(image, caption="Uploaded Image", use_column_width=True)
|
| 175 |
-
|
| 176 |
-
# Process image
|
| 177 |
-
if use_face_detection:
|
| 178 |
-
faces, opencv_image = detect_faces(image)
|
| 179 |
-
|
| 180 |
-
if len(faces) > 0:
|
| 181 |
-
emotions = []
|
| 182 |
-
for face in faces:
|
| 183 |
-
face_img = process_image_for_emotion(image, face)
|
| 184 |
-
emotions.append(predict_emotion(face_img))
|
| 185 |
-
|
| 186 |
-
# Draw faces with emotions
|
| 187 |
-
result_image = draw_faces_with_emotions(opencv_image, faces, emotions)
|
| 188 |
-
|
| 189 |
-
with col2:
|
| 190 |
-
st.image(result_image, caption="Detected Emotions", channels="BGR", use_column_width=True)
|
| 191 |
-
|
| 192 |
-
# Display predictions
|
| 193 |
-
st.subheader("Detected Emotions:")
|
| 194 |
-
for i, (emotion, face) in enumerate(zip(emotions, faces)):
|
| 195 |
-
if emotion["score"] >= confidence_threshold:
|
| 196 |
-
st.write(f"Face {i+1}: **{emotion['label']}** (Confidence: {emotion['score']:.2f})")
|
| 197 |
-
|
| 198 |
-
# Show confidence bars
|
| 199 |
-
top_emotions = classifier(process_image_for_emotion(image, face))
|
| 200 |
-
labels = [item["label"] for item in top_emotions]
|
| 201 |
-
scores = [item["score"] for item in top_emotions]
|
| 202 |
-
|
| 203 |
-
fig = go.Figure(go.Bar(
|
| 204 |
-
x=scores,
|
| 205 |
-
y=labels,
|
| 206 |
-
orientation='h'
|
| 207 |
-
))
|
| 208 |
-
fig.update_layout(
|
| 209 |
-
title=f"Emotion Confidence - Face {i+1}",
|
| 210 |
-
xaxis_title="Confidence",
|
| 211 |
-
yaxis_title="Emotion",
|
| 212 |
-
height=300
|
| 213 |
-
)
|
| 214 |
-
st.plotly_chart(fig, use_container_width=True)
|
| 215 |
-
else:
|
| 216 |
-
st.warning("No faces detected in the image. Try another image or disable face detection.")
|
| 217 |
-
else:
|
| 218 |
-
# Process the whole image
|
| 219 |
-
prediction = predict_emotion(image)
|
| 220 |
-
st.subheader("Prediction:")
|
| 221 |
-
st.write(f"**Emotion:** {prediction['label']}")
|
| 222 |
-
st.write(f"**Confidence:** {prediction['score']:.2f}")
|
| 223 |
-
|
| 224 |
-
elif input_method == "Capture Image":
|
| 225 |
-
picture = st.camera_input("Capture an Image")
|
| 226 |
|
| 227 |
-
|
| 228 |
-
|
| 229 |
-
|
| 230 |
-
col1, col2 = st.columns(2)
|
| 231 |
-
with col1:
|
| 232 |
-
st.image(image, caption="Captured Image", use_column_width=True)
|
| 233 |
-
|
| 234 |
-
# Process image
|
| 235 |
-
if use_face_detection:
|
| 236 |
-
faces, opencv_image = detect_faces(image)
|
| 237 |
-
|
| 238 |
-
if len(faces) > 0:
|
| 239 |
-
emotions = []
|
| 240 |
-
for face in faces:
|
| 241 |
-
face_img = process_image_for_emotion(image, face)
|
| 242 |
-
emotions.append(predict_emotion(face_img))
|
| 243 |
-
|
| 244 |
-
# Draw faces with emotions
|
| 245 |
-
result_image = draw_faces_with_emotions(opencv_image, faces, emotions)
|
| 246 |
-
|
| 247 |
-
with col2:
|
| 248 |
-
st.image(result_image, caption="Detected Emotions", channels="BGR", use_column_width=True)
|
| 249 |
-
|
| 250 |
-
# Display predictions
|
| 251 |
-
st.subheader("Detected Emotions:")
|
| 252 |
-
for i, (emotion, face) in enumerate(zip(emotions, faces)):
|
| 253 |
-
if emotion["score"] >= confidence_threshold:
|
| 254 |
-
st.write(f"Face {i+1}: **{emotion['label']}** (Confidence: {emotion['score']:.2f})")
|
| 255 |
-
else:
|
| 256 |
-
st.warning("No faces detected in the image. Try another image or disable face detection.")
|
| 257 |
-
else:
|
| 258 |
-
# Process the whole image
|
| 259 |
-
prediction = predict_emotion(image)
|
| 260 |
-
st.subheader("Prediction:")
|
| 261 |
-
st.write(f"**Emotion:** {prediction['label']}")
|
| 262 |
-
st.write(f"**Confidence:** {prediction['score']:.2f}")
|
| 263 |
-
|
| 264 |
-
elif input_method == "Real-time Webcam":
|
| 265 |
-
st.subheader("Real-time Emotion Detection")
|
| 266 |
-
st.write("Click 'Start' to begin real-time emotion detection using your webcam.")
|
| 267 |
|
| 268 |
-
|
| 269 |
-
|
| 270 |
|
| 271 |
-
#
|
| 272 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 273 |
|
| 274 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 275 |
chart_placeholder = st.empty()
|
| 276 |
|
| 277 |
# Initialize session state for tracking emotions over time
|
|
@@ -279,6 +254,7 @@ elif input_method == "Real-time Webcam":
|
|
| 279 |
st.session_state.emotion_history = {}
|
| 280 |
st.session_state.last_update_time = time.time()
|
| 281 |
st.session_state.frame_count = 0
|
|
|
|
| 282 |
|
| 283 |
# Start/Stop button
|
| 284 |
start_button = st.button("Start" if 'running' not in st.session_state or not st.session_state.running else "Stop")
|
|
@@ -286,171 +262,196 @@ elif input_method == "Real-time Webcam":
|
|
| 286 |
if start_button:
|
| 287 |
st.session_state.running = not st.session_state.get('running', False)
|
| 288 |
|
| 289 |
-
# If running, capture and process
|
| 290 |
if st.session_state.get('running', False):
|
| 291 |
try:
|
| 292 |
-
#
|
| 293 |
-
|
| 294 |
-
|
| 295 |
-
|
| 296 |
-
if not cap.isOpened():
|
| 297 |
-
st.error("Could not open webcam. Please check your camera settings.")
|
| 298 |
-
st.session_state.running = False
|
| 299 |
else:
|
| 300 |
-
|
| 301 |
-
emotion_deques = {}
|
| 302 |
-
timestamp_deque = deque(maxlen=30*history_length) # Store timestamps for X seconds at 30fps
|
| 303 |
|
| 304 |
-
|
| 305 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 306 |
ret, frame = cap.read()
|
| 307 |
-
|
| 308 |
-
|
| 309 |
-
|
| 310 |
-
|
| 311 |
-
|
| 312 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 313 |
frame = cv2.flip(frame, 1)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
|
| 315 |
-
|
| 316 |
-
|
| 317 |
-
|
| 318 |
-
|
| 319 |
-
|
| 320 |
-
|
| 321 |
|
| 322 |
-
|
| 323 |
-
|
| 324 |
-
emotions = []
|
| 325 |
-
for face in faces:
|
| 326 |
-
face_img = process_image_for_emotion(frame, face)
|
| 327 |
-
emotions.append(predict_emotion(face_img))
|
| 328 |
-
|
| 329 |
-
# Draw faces with emotions
|
| 330 |
-
frame = draw_faces_with_emotions(frame, faces, emotions)
|
| 331 |
-
|
| 332 |
-
# Update emotion history
|
| 333 |
-
current_time = time.time()
|
| 334 |
-
timestamp_deque.append(current_time)
|
| 335 |
-
|
| 336 |
-
for i, emotion in enumerate(emotions):
|
| 337 |
-
if emotion["score"] >= confidence_threshold:
|
| 338 |
-
face_id = f"Face {i+1}"
|
| 339 |
-
if face_id not in emotion_deques:
|
| 340 |
-
emotion_deques[face_id] = deque(maxlen=30*history_length)
|
| 341 |
-
|
| 342 |
-
emotion_deques[face_id].append({
|
| 343 |
-
"emotion": emotion["label"],
|
| 344 |
-
"confidence": emotion["score"],
|
| 345 |
-
"time": current_time
|
| 346 |
-
})
|
| 347 |
-
else:
|
| 348 |
-
# No faces detected
|
| 349 |
-
pass
|
| 350 |
-
else:
|
| 351 |
-
# Process the whole frame
|
| 352 |
-
pil_image = Image.fromarray(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
|
| 353 |
-
emotion = predict_emotion(pil_image)
|
| 354 |
-
|
| 355 |
-
# Display emotion on frame
|
| 356 |
-
cv2.putText(
|
| 357 |
-
frame,
|
| 358 |
-
f"{emotion['label']} ({emotion['score']:.2f})",
|
| 359 |
-
(10, 30),
|
| 360 |
-
cv2.FONT_HERSHEY_SIMPLEX,
|
| 361 |
-
1,
|
| 362 |
-
(0, 255, 0),
|
| 363 |
-
2
|
| 364 |
-
)
|
| 365 |
|
| 366 |
# Update emotion history
|
| 367 |
current_time = time.time()
|
| 368 |
timestamp_deque.append(current_time)
|
| 369 |
|
| 370 |
-
|
| 371 |
-
|
| 372 |
-
|
| 373 |
-
|
| 374 |
-
|
| 375 |
-
|
| 376 |
-
|
| 377 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 378 |
|
| 379 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 380 |
current_time = time.time()
|
| 381 |
-
|
| 382 |
-
if time_diff >= 1.0: # Update every second
|
| 383 |
-
fps = st.session_state.frame_count / time_diff
|
| 384 |
-
st.session_state.last_update_time = current_time
|
| 385 |
-
st.session_state.frame_count = 0
|
| 386 |
-
|
| 387 |
-
# Update metrics
|
| 388 |
-
with metrics_placeholder.container():
|
| 389 |
-
cols = st.columns(3)
|
| 390 |
-
cols[0].metric("FPS", f"{fps:.1f}")
|
| 391 |
-
cols[1].metric("Faces Detected", len(faces) if use_face_detection else "N/A")
|
| 392 |
|
| 393 |
-
|
| 394 |
-
|
| 395 |
|
| 396 |
-
|
| 397 |
-
|
| 398 |
-
|
| 399 |
-
|
| 400 |
-
|
| 401 |
-
|
| 402 |
-
|
| 403 |
-
|
| 404 |
-
|
| 405 |
-
|
| 406 |
-
|
| 407 |
-
|
| 408 |
-
|
| 409 |
-
|
| 410 |
-
|
| 411 |
-
|
| 412 |
-
|
| 413 |
-
|
| 414 |
-
|
| 415 |
-
fig = go.Figure(data=[go.Pie(
|
| 416 |
-
labels=list(emotion_counts.keys()),
|
| 417 |
-
values=list(emotion_counts.values()),
|
| 418 |
-
hole=.3
|
| 419 |
-
)])
|
| 420 |
-
fig.update_layout(title=f"Emotion Distribution - {face_id}")
|
| 421 |
-
st.plotly_chart(fig, use_container_width=True)
|
| 422 |
-
|
| 423 |
-
# Create line chart for emotion confidence over time
|
| 424 |
-
emotions = list(emotion_data)[-20:] # Get the last 20 entries
|
| 425 |
-
times = [(e["time"] - emotions[0]["time"]) for e in emotions]
|
| 426 |
-
confidences = [e["confidence"] for e in emotions]
|
| 427 |
-
emotion_labels = [e["emotion"] for e in emotions]
|
| 428 |
-
|
| 429 |
-
fig = go.Figure()
|
| 430 |
-
fig.add_trace(go.Scatter(
|
| 431 |
-
x=times,
|
| 432 |
-
y=confidences,
|
| 433 |
-
mode='lines+markers',
|
| 434 |
-
text=emotion_labels,
|
| 435 |
-
hoverinfo='text+y'
|
| 436 |
-
))
|
| 437 |
-
fig.update_layout(
|
| 438 |
-
title=f"Emotion Confidence Over Time - {face_id}",
|
| 439 |
-
xaxis_title="Time (seconds)",
|
| 440 |
-
yaxis_title="Confidence",
|
| 441 |
-
yaxis=dict(range=[0, 1])
|
| 442 |
-
)
|
| 443 |
-
st.plotly_chart(fig, use_container_width=True)
|
| 444 |
-
else:
|
| 445 |
-
st.info(f"No emotion data available for {face_id} yet.")
|
| 446 |
-
else:
|
| 447 |
-
st.info("No emotion data available yet.")
|
| 448 |
|
| 449 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 450 |
cap.release()
|
| 451 |
|
| 452 |
except Exception as e:
|
| 453 |
-
st.error(f"Error during
|
| 454 |
st.session_state.running = False
|
| 455 |
else:
|
| 456 |
# Display a placeholder image when not running
|
|
@@ -466,12 +467,128 @@ elif input_method == "Real-time Webcam":
|
|
| 466 |
)
|
| 467 |
video_placeholder.image(placeholder_img, channels="BGR", use_column_width=True)
|
| 468 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 469 |
# --- Footer ---
|
| 470 |
st.markdown("---")
|
| 471 |
st.markdown("""
|
| 472 |
**Tips for Best Results:**
|
|
|
|
| 473 |
- Ensure good lighting for accurate face detection
|
| 474 |
-
- Position
|
| 475 |
- Try different emotion models for comparison
|
| 476 |
-
- Adjust the confidence threshold if emotions aren't being detected
|
| 477 |
""")
|
|
|
|
| 7 |
from PIL import Image
|
| 8 |
import torch
|
| 9 |
from collections import deque
|
| 10 |
+
import os
|
| 11 |
+
import tempfile
|
| 12 |
|
| 13 |
# Set page config
|
| 14 |
st.set_page_config(
|
|
|
|
| 20 |
# --- App Title and Description ---
|
| 21 |
st.title("Advanced Real-Time Emotion Detection")
|
| 22 |
st.write("""
|
| 23 |
+
This app detects emotions in real-time using webcam, video files, or images.
|
| 24 |
+
If your webcam isn't working, try the simulation mode or upload a video file.
|
| 25 |
""")
|
| 26 |
|
| 27 |
# --- Load Models ---
|
|
|
|
| 53 |
list(model_options.keys())
|
| 54 |
)
|
| 55 |
|
| 56 |
+
# Input method selection with addition of video upload and simulation
|
| 57 |
input_method = st.sidebar.radio(
|
| 58 |
"Choose Input Method",
|
| 59 |
+
["Real-time Webcam", "Upload Video", "Simulation Mode", "Upload an Image", "Capture Image"]
|
| 60 |
)
|
| 61 |
|
| 62 |
# Confidence threshold
|
|
|
|
| 71 |
# Face detection toggle
|
| 72 |
use_face_detection = st.sidebar.checkbox("Enable Face Detection", value=True)
|
| 73 |
|
| 74 |
+
# Processing speed for video/simulation
|
| 75 |
+
if input_method in ["Upload Video", "Simulation Mode"]:
|
| 76 |
+
processing_speed = st.sidebar.slider(
|
| 77 |
+
"Processing Speed",
|
| 78 |
+
min_value=0.1,
|
| 79 |
+
max_value=2.0,
|
| 80 |
+
value=1.0,
|
| 81 |
+
step=0.1,
|
| 82 |
+
help="Adjust the speed of video processing (higher is faster)"
|
| 83 |
+
)
|
| 84 |
+
|
| 85 |
# History length for real-time tracking
|
| 86 |
+
if input_method in ["Real-time Webcam", "Upload Video", "Simulation Mode"]:
|
| 87 |
history_length = st.sidebar.slider(
|
| 88 |
"Emotion History Length (seconds)",
|
| 89 |
min_value=5,
|
|
|
|
| 174 |
|
| 175 |
return img
|
| 176 |
|
| 177 |
+
def generate_simulated_face(frame_num, canvas_size=(640, 480)):
|
| 178 |
+
"""Generate a simulated face with changing expressions."""
|
| 179 |
+
# Create a blank canvas
|
| 180 |
+
canvas = np.ones((canvas_size[1], canvas_size[0], 3), dtype=np.uint8) * 230
|
| 181 |
|
| 182 |
+
# Calculate center position and face size
|
| 183 |
+
center_x, center_y = canvas_size[0] // 2, canvas_size[1] // 2
|
| 184 |
+
face_radius = min(canvas_size) // 4
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 185 |
|
| 186 |
+
# Face movement based on frame number
|
| 187 |
+
movement_x = int(np.sin(frame_num * 0.02) * 50)
|
| 188 |
+
movement_y = int(np.cos(frame_num * 0.03) * 30)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 189 |
|
| 190 |
+
face_x = center_x + movement_x
|
| 191 |
+
face_y = center_y + movement_y
|
| 192 |
|
| 193 |
+
# Draw face circle
|
| 194 |
+
cv2.circle(canvas, (face_x, face_y), face_radius, (220, 210, 180), -1)
|
| 195 |
+
|
| 196 |
+
# Draw eyes
|
| 197 |
+
eye_y = face_y - int(face_radius * 0.2)
|
| 198 |
+
left_eye_x = face_x - int(face_radius * 0.5)
|
| 199 |
+
right_eye_x = face_x + int(face_radius * 0.5)
|
| 200 |
+
eye_size = max(5, face_radius // 8)
|
| 201 |
+
|
| 202 |
+
# Blink occasionally
|
| 203 |
+
if frame_num % 50 > 45: # Blink every 50 frames for 5 frames
|
| 204 |
+
cv2.ellipse(canvas, (left_eye_x, eye_y), (eye_size, 1), 0, 0, 360, (30, 30, 30), -1)
|
| 205 |
+
cv2.ellipse(canvas, (right_eye_x, eye_y), (eye_size, 1), 0, 0, 360, (30, 30, 30), -1)
|
| 206 |
+
else:
|
| 207 |
+
cv2.circle(canvas, (left_eye_x, eye_y), eye_size, (255, 255, 255), -1)
|
| 208 |
+
cv2.circle(canvas, (right_eye_x, eye_y), eye_size, (255, 255, 255), -1)
|
| 209 |
+
cv2.circle(canvas, (left_eye_x, eye_y), eye_size-2, (70, 70, 70), -1)
|
| 210 |
+
cv2.circle(canvas, (right_eye_x, eye_y), eye_size-2, (70, 70, 70), -1)
|
| 211 |
+
|
| 212 |
+
# Draw mouth - change shape based on frame number to simulate different emotions
|
| 213 |
+
mouth_y = face_y + int(face_radius * 0.3)
|
| 214 |
+
mouth_width = int(face_radius * 0.6)
|
| 215 |
+
mouth_height = int(face_radius * 0.2)
|
| 216 |
+
|
| 217 |
+
# Cycle through different emotions based on frame number
|
| 218 |
+
emotion_cycle = (frame_num // 100) % 4
|
| 219 |
+
|
| 220 |
+
if emotion_cycle == 0: # Happy
|
| 221 |
+
# Smile
|
| 222 |
+
cv2.ellipse(canvas, (face_x, mouth_y), (mouth_width, mouth_height),
|
| 223 |
+
0, 0, 180, (50, 50, 50), 2)
|
| 224 |
+
elif emotion_cycle == 1: # Sad
|
| 225 |
+
# Frown
|
| 226 |
+
cv2.ellipse(canvas, (face_x, mouth_y + mouth_height),
|
| 227 |
+
(mouth_width, mouth_height), 0, 180, 360, (50, 50, 50), 2)
|
| 228 |
+
elif emotion_cycle == 2: # Surprised
|
| 229 |
+
# O mouth
|
| 230 |
+
cv2.circle(canvas, (face_x, mouth_y), mouth_height, (50, 50, 50), 2)
|
| 231 |
+
else: # Neutral
|
| 232 |
+
# Straight line
|
| 233 |
+
cv2.line(canvas, (face_x - mouth_width//2, mouth_y),
|
| 234 |
+
(face_x + mouth_width//2, mouth_y), (50, 50, 50), 2)
|
| 235 |
+
|
| 236 |
+
# Add some text showing what emotion is being simulated
|
| 237 |
+
emotions = ["Happy", "Sad", "Surprised", "Neutral"]
|
| 238 |
+
cv2.putText(canvas, f"Simulating: {emotions[emotion_cycle]}",
|
| 239 |
+
(20, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (50, 50, 50), 2)
|
| 240 |
+
cv2.putText(canvas, "Simulation Mode - No webcam required",
|
| 241 |
+
(20, canvas_size[1] - 20), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (100, 100, 100), 1)
|
| 242 |
|
| 243 |
+
return canvas
|
| 244 |
+
|
| 245 |
+
def process_video_feed(feed_source, is_simulation=False):
|
| 246 |
+
"""Process video feed (webcam, video file, or simulation)."""
|
| 247 |
+
# Create placeholders
|
| 248 |
+
video_placeholder = st.empty()
|
| 249 |
+
metrics_placeholder = st.empty()
|
| 250 |
chart_placeholder = st.empty()
|
| 251 |
|
| 252 |
# Initialize session state for tracking emotions over time
|
|
|
|
| 254 |
st.session_state.emotion_history = {}
|
| 255 |
st.session_state.last_update_time = time.time()
|
| 256 |
st.session_state.frame_count = 0
|
| 257 |
+
st.session_state.simulation_frame = 0
|
| 258 |
|
| 259 |
# Start/Stop button
|
| 260 |
start_button = st.button("Start" if 'running' not in st.session_state or not st.session_state.running else "Stop")
|
|
|
|
| 262 |
if start_button:
|
| 263 |
st.session_state.running = not st.session_state.get('running', False)
|
| 264 |
|
| 265 |
+
# If running, capture and process video feed
|
| 266 |
if st.session_state.get('running', False):
|
| 267 |
try:
|
| 268 |
+
# Initialize video source
|
| 269 |
+
if is_simulation:
|
| 270 |
+
# No need to open a video source for simulation
|
| 271 |
+
pass
|
|
|
|
|
|
|
|
|
|
| 272 |
else:
|
| 273 |
+
cap = feed_source
|
|
|
|
|
|
|
| 274 |
|
| 275 |
+
# Check if video source opened successfully
|
| 276 |
+
if not cap.isOpened():
|
| 277 |
+
st.error("Could not open video source. Please check your settings.")
|
| 278 |
+
st.session_state.running = False
|
| 279 |
+
return
|
| 280 |
+
|
| 281 |
+
# Create deques for tracking emotions
|
| 282 |
+
emotion_deques = {}
|
| 283 |
+
timestamp_deque = deque(maxlen=30*history_length) # Store timestamps for X seconds at 30fps
|
| 284 |
+
|
| 285 |
+
while st.session_state.get('running', False):
|
| 286 |
+
# Get frame
|
| 287 |
+
if is_simulation:
|
| 288 |
+
# Generate a simulated frame
|
| 289 |
+
frame = generate_simulated_face(st.session_state.simulation_frame)
|
| 290 |
+
st.session_state.simulation_frame += 1
|
| 291 |
+
ret = True
|
| 292 |
+
else:
|
| 293 |
+
# Read from video source
|
| 294 |
ret, frame = cap.read()
|
| 295 |
+
|
| 296 |
+
if not ret:
|
| 297 |
+
if is_simulation:
|
| 298 |
+
st.error("Simulation error")
|
| 299 |
+
elif input_method == "Upload Video":
|
| 300 |
+
# For video files, loop back to the beginning
|
| 301 |
+
cap.set(cv2.CAP_PROP_POS_FRAMES, 0)
|
| 302 |
+
continue
|
| 303 |
+
else:
|
| 304 |
+
st.error("Failed to capture frame from video source")
|
| 305 |
+
break
|
| 306 |
+
|
| 307 |
+
# For webcam, flip horizontally for a more natural view
|
| 308 |
+
if input_method == "Real-time Webcam" and not is_simulation:
|
| 309 |
frame = cv2.flip(frame, 1)
|
| 310 |
+
|
| 311 |
+
# Increment frame count for FPS calculation
|
| 312 |
+
st.session_state.frame_count += 1
|
| 313 |
+
|
| 314 |
+
# Detect faces
|
| 315 |
+
if use_face_detection:
|
| 316 |
+
faces, _ = detect_faces(frame)
|
| 317 |
|
| 318 |
+
if len(faces) > 0:
|
| 319 |
+
# Process each face
|
| 320 |
+
emotions = []
|
| 321 |
+
for face in faces:
|
| 322 |
+
face_img = process_image_for_emotion(frame, face)
|
| 323 |
+
emotions.append(predict_emotion(face_img))
|
| 324 |
|
| 325 |
+
# Draw faces with emotions
|
| 326 |
+
frame = draw_faces_with_emotions(frame, faces, emotions)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 327 |
|
| 328 |
# Update emotion history
|
| 329 |
current_time = time.time()
|
| 330 |
timestamp_deque.append(current_time)
|
| 331 |
|
| 332 |
+
for i, emotion in enumerate(emotions):
|
| 333 |
+
if emotion["score"] >= confidence_threshold:
|
| 334 |
+
face_id = f"Face {i+1}"
|
| 335 |
+
if face_id not in emotion_deques:
|
| 336 |
+
emotion_deques[face_id] = deque(maxlen=30*history_length)
|
| 337 |
+
|
| 338 |
+
emotion_deques[face_id].append({
|
| 339 |
+
"emotion": emotion["label"],
|
| 340 |
+
"confidence": emotion["score"],
|
| 341 |
+
"time": current_time
|
| 342 |
+
})
|
| 343 |
+
else:
|
| 344 |
+
# No faces detected
|
| 345 |
+
pass
|
| 346 |
+
else:
|
| 347 |
+
# Process the whole frame
|
| 348 |
+
pil_image = Image.fromarray(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
|
| 349 |
+
emotion = predict_emotion(pil_image)
|
| 350 |
|
| 351 |
+
# Display emotion on frame
|
| 352 |
+
cv2.putText(
|
| 353 |
+
frame,
|
| 354 |
+
f"{emotion['label']} ({emotion['score']:.2f})",
|
| 355 |
+
(10, 30),
|
| 356 |
+
cv2.FONT_HERSHEY_SIMPLEX,
|
| 357 |
+
1,
|
| 358 |
+
(0, 255, 0),
|
| 359 |
+
2
|
| 360 |
+
)
|
| 361 |
+
|
| 362 |
+
# Update emotion history
|
| 363 |
current_time = time.time()
|
| 364 |
+
timestamp_deque.append(current_time)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 365 |
|
| 366 |
+
if "Frame" not in emotion_deques:
|
| 367 |
+
emotion_deques["Frame"] = deque(maxlen=30*history_length)
|
| 368 |
|
| 369 |
+
emotion_deques["Frame"].append({
|
| 370 |
+
"emotion": emotion["label"],
|
| 371 |
+
"confidence": emotion["score"],
|
| 372 |
+
"time": current_time
|
| 373 |
+
})
|
| 374 |
+
|
| 375 |
+
# Calculate FPS
|
| 376 |
+
current_time = time.time()
|
| 377 |
+
time_diff = current_time - st.session_state.last_update_time
|
| 378 |
+
if time_diff >= 1.0: # Update every second
|
| 379 |
+
fps = st.session_state.frame_count / time_diff
|
| 380 |
+
st.session_state.last_update_time = current_time
|
| 381 |
+
st.session_state.frame_count = 0
|
| 382 |
+
|
| 383 |
+
# Update metrics
|
| 384 |
+
with metrics_placeholder.container():
|
| 385 |
+
cols = st.columns(3)
|
| 386 |
+
cols[0].metric("FPS", f"{fps:.1f}")
|
| 387 |
+
cols[1].metric("Faces Detected", len(faces) if use_face_detection else "N/A")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 388 |
|
| 389 |
+
# Display the frame
|
| 390 |
+
video_placeholder.image(frame, channels="BGR", use_column_width=True)
|
| 391 |
+
|
| 392 |
+
# Update emotion history chart periodically
|
| 393 |
+
if len(timestamp_deque) > 0 and time_diff >= 0.5: # Update chart every 0.5 seconds
|
| 394 |
+
with chart_placeholder.container():
|
| 395 |
+
# Create tabs for each face
|
| 396 |
+
if len(emotion_deques) > 0:
|
| 397 |
+
tabs = st.tabs(list(emotion_deques.keys()))
|
| 398 |
+
|
| 399 |
+
for i, (face_id, emotion_data) in enumerate(emotion_deques.items()):
|
| 400 |
+
with tabs[i]:
|
| 401 |
+
if len(emotion_data) > 0:
|
| 402 |
+
# Count occurrences of each emotion
|
| 403 |
+
emotion_counts = {}
|
| 404 |
+
for entry in emotion_data:
|
| 405 |
+
emotion = entry["emotion"]
|
| 406 |
+
if emotion not in emotion_counts:
|
| 407 |
+
emotion_counts[emotion] = 0
|
| 408 |
+
emotion_counts[emotion] += 1
|
| 409 |
+
|
| 410 |
+
# Create pie chart for emotion distribution
|
| 411 |
+
fig = go.Figure(data=[go.Pie(
|
| 412 |
+
labels=list(emotion_counts.keys()),
|
| 413 |
+
values=list(emotion_counts.values()),
|
| 414 |
+
hole=.3
|
| 415 |
+
)])
|
| 416 |
+
fig.update_layout(title=f"Emotion Distribution - {face_id}")
|
| 417 |
+
st.plotly_chart(fig, use_container_width=True)
|
| 418 |
+
|
| 419 |
+
# Create line chart for emotion confidence over time
|
| 420 |
+
emotions = list(emotion_data)[-20:] # Get the last 20 entries
|
| 421 |
+
times = [(e["time"] - emotions[0]["time"]) for e in emotions]
|
| 422 |
+
confidences = [e["confidence"] for e in emotions]
|
| 423 |
+
emotion_labels = [e["emotion"] for e in emotions]
|
| 424 |
+
|
| 425 |
+
fig = go.Figure()
|
| 426 |
+
fig.add_trace(go.Scatter(
|
| 427 |
+
x=times,
|
| 428 |
+
y=confidences,
|
| 429 |
+
mode='lines+markers',
|
| 430 |
+
text=emotion_labels,
|
| 431 |
+
hoverinfo='text+y'
|
| 432 |
+
))
|
| 433 |
+
fig.update_layout(
|
| 434 |
+
title=f"Emotion Confidence Over Time - {face_id}",
|
| 435 |
+
xaxis_title="Time (seconds)",
|
| 436 |
+
yaxis_title="Confidence",
|
| 437 |
+
yaxis=dict(range=[0, 1])
|
| 438 |
+
)
|
| 439 |
+
st.plotly_chart(fig, use_container_width=True)
|
| 440 |
+
else:
|
| 441 |
+
st.info(f"No emotion data available for {face_id} yet.")
|
| 442 |
+
else:
|
| 443 |
+
st.info("No emotion data available yet.")
|
| 444 |
+
|
| 445 |
+
# Control processing speed for videos and simulation
|
| 446 |
+
if input_method in ["Upload Video", "Simulation Mode"]:
|
| 447 |
+
time.sleep(0.03 / processing_speed) # Adjust delay based on processing_speed
|
| 448 |
+
|
| 449 |
+
# Release resources when done
|
| 450 |
+
if not is_simulation and cap.isOpened():
|
| 451 |
cap.release()
|
| 452 |
|
| 453 |
except Exception as e:
|
| 454 |
+
st.error(f"Error during processing: {str(e)}")
|
| 455 |
st.session_state.running = False
|
| 456 |
else:
|
| 457 |
# Display a placeholder image when not running
|
|
|
|
| 467 |
)
|
| 468 |
video_placeholder.image(placeholder_img, channels="BGR", use_column_width=True)
|
| 469 |
|
| 470 |
+
# --- Process uploaded image ---
|
| 471 |
+
def process_static_image(image):
|
| 472 |
+
col1, col2 = st.columns(2)
|
| 473 |
+
with col1:
|
| 474 |
+
st.image(image, caption="Image", use_column_width=True)
|
| 475 |
+
|
| 476 |
+
# Process image
|
| 477 |
+
if use_face_detection:
|
| 478 |
+
faces, opencv_image = detect_faces(image)
|
| 479 |
+
|
| 480 |
+
if len(faces) > 0:
|
| 481 |
+
emotions = []
|
| 482 |
+
for face in faces:
|
| 483 |
+
face_img = process_image_for_emotion(image, face)
|
| 484 |
+
emotions.append(predict_emotion(face_img))
|
| 485 |
+
|
| 486 |
+
# Draw faces with emotions
|
| 487 |
+
result_image = draw_faces_with_emotions(opencv_image, faces, emotions)
|
| 488 |
+
|
| 489 |
+
with col2:
|
| 490 |
+
st.image(result_image, caption="Detected Emotions", channels="BGR", use_column_width=True)
|
| 491 |
+
|
| 492 |
+
# Display predictions
|
| 493 |
+
st.subheader("Detected Emotions:")
|
| 494 |
+
for i, (emotion, face) in enumerate(zip(emotions, faces)):
|
| 495 |
+
if emotion["score"] >= confidence_threshold:
|
| 496 |
+
st.write(f"Face {i+1}: **{emotion['label']}** (Confidence: {emotion['score']:.2f})")
|
| 497 |
+
|
| 498 |
+
# Show confidence bars
|
| 499 |
+
top_emotions = classifier(process_image_for_emotion(image, face))
|
| 500 |
+
labels = [item["label"] for item in top_emotions]
|
| 501 |
+
scores = [item["score"] for item in top_emotions]
|
| 502 |
+
|
| 503 |
+
fig = go.Figure(go.Bar(
|
| 504 |
+
x=scores,
|
| 505 |
+
y=labels,
|
| 506 |
+
orientation='h'
|
| 507 |
+
))
|
| 508 |
+
fig.update_layout(
|
| 509 |
+
title=f"Emotion Confidence - Face {i+1}",
|
| 510 |
+
xaxis_title="Confidence",
|
| 511 |
+
yaxis_title="Emotion",
|
| 512 |
+
height=300
|
| 513 |
+
)
|
| 514 |
+
st.plotly_chart(fig, use_container_width=True)
|
| 515 |
+
else:
|
| 516 |
+
st.warning("No faces detected in the image. Try another image or disable face detection.")
|
| 517 |
+
else:
|
| 518 |
+
# Process the whole image
|
| 519 |
+
prediction = predict_emotion(image)
|
| 520 |
+
st.subheader("Prediction:")
|
| 521 |
+
st.write(f"**Emotion:** {prediction['label']}")
|
| 522 |
+
st.write(f"**Confidence:** {prediction['score']:.2f}")
|
| 523 |
+
|
| 524 |
+
# --- Main App Logic ---
|
| 525 |
+
if input_method == "Upload an Image":
|
| 526 |
+
uploaded_file = st.file_uploader("Choose an image file", type=["jpg", "jpeg", "png"])
|
| 527 |
+
|
| 528 |
+
if uploaded_file is not None:
|
| 529 |
+
# Load and display image
|
| 530 |
+
image = Image.open(uploaded_file).convert("RGB")
|
| 531 |
+
process_static_image(image)
|
| 532 |
+
|
| 533 |
+
elif input_method == "Capture Image":
|
| 534 |
+
picture = st.camera_input("Capture an Image")
|
| 535 |
+
|
| 536 |
+
if picture is not None:
|
| 537 |
+
image = Image.open(picture).convert("RGB")
|
| 538 |
+
process_static_image(image)
|
| 539 |
+
|
| 540 |
+
elif input_method == "Upload Video":
|
| 541 |
+
uploaded_video = st.file_uploader("Upload a video file", type=["mp4", "avi", "mov", "mkv"])
|
| 542 |
+
|
| 543 |
+
if uploaded_video is not None:
|
| 544 |
+
# Save the uploaded video to a temporary file
|
| 545 |
+
tfile = tempfile.NamedTemporaryFile(delete=False)
|
| 546 |
+
tfile.write(uploaded_video.read())
|
| 547 |
+
|
| 548 |
+
# Open the video file
|
| 549 |
+
cap = cv2.VideoCapture(tfile.name)
|
| 550 |
+
|
| 551 |
+
# Process the video
|
| 552 |
+
process_video_feed(cap)
|
| 553 |
+
|
| 554 |
+
# Clean up the temporary file
|
| 555 |
+
os.unlink(tfile.name)
|
| 556 |
+
|
| 557 |
+
elif input_method == "Simulation Mode":
|
| 558 |
+
st.info("Simulation mode uses a generated animated face. No webcam required!")
|
| 559 |
+
process_video_feed(None, is_simulation=True)
|
| 560 |
+
|
| 561 |
+
elif input_method == "Real-time Webcam":
|
| 562 |
+
try:
|
| 563 |
+
# First check if we can access the webcam
|
| 564 |
+
cap = cv2.VideoCapture(0)
|
| 565 |
+
if not cap.isOpened():
|
| 566 |
+
st.error("Could not open webcam. Please try the Simulation Mode instead.")
|
| 567 |
+
st.info("If you're using Streamlit in a browser, make sure you've granted camera permissions.")
|
| 568 |
+
|
| 569 |
+
# Show troubleshooting tips
|
| 570 |
+
with st.expander("Webcam Troubleshooting Tips"):
|
| 571 |
+
st.markdown("""
|
| 572 |
+
1. **Check Browser Permissions**: Make sure your browser has permission to access your camera.
|
| 573 |
+
2. **Close Other Applications**: Other applications might be using your webcam.
|
| 574 |
+
3. **Refresh the Page**: Sometimes simply refreshing can resolve the issue.
|
| 575 |
+
4. **Try a Different Browser**: Some browsers handle webcam access better than others.
|
| 576 |
+
5. **Use Simulation Mode**: If you cannot get the webcam working, use the Simulation Mode.
|
| 577 |
+
""")
|
| 578 |
+
else:
|
| 579 |
+
# Webcam available, process it
|
| 580 |
+
process_video_feed(cap)
|
| 581 |
+
except Exception as e:
|
| 582 |
+
st.error(f"Error accessing webcam: {str(e)}")
|
| 583 |
+
st.info("Please try the Simulation Mode instead, which doesn't require webcam access.")
|
| 584 |
+
|
| 585 |
# --- Footer ---
|
| 586 |
st.markdown("---")
|
| 587 |
st.markdown("""
|
| 588 |
**Tips for Best Results:**
|
| 589 |
+
- If webcam doesn't work, try "Simulation Mode" or "Upload Video" options
|
| 590 |
- Ensure good lighting for accurate face detection
|
| 591 |
+
- Position faces clearly in the frame
|
| 592 |
- Try different emotion models for comparison
|
| 593 |
+
- Adjust the confidence threshold if emotions aren't being detected correctly
|
| 594 |
""")
|