Spaces:
Sleeping
Sleeping
File size: 3,937 Bytes
687c7d2 833d8e7 687c7d2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
from streamlit_webrtc import webrtc_streamer, WebRtcMode, RTCConfiguration
import streamlit as st
import av
import cv2
import numpy as np
from deepface import DeepFace
import time
st.set_page_config(page_title="AI Facial Interview Monitor", layout="wide")
st.title(":blue[MOCKVIEWER - Face Monitoring System]")
# Load Haar Cascade models
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + "haarcascade_frontalface_default.xml")
eye_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + "haarcascade_eye.xml")
# Upload reference image
uploaded_image = st.file_uploader("Upload your reference face image", type=["jpg", "jpeg", "png"])
if uploaded_image:
user_ref_img = cv2.imdecode(np.frombuffer(uploaded_image.read(), np.uint8), cv2.IMREAD_COLOR)
st.image(user_ref_img, caption="Reference Image", use_column_width=True)
face_detected_time = time.time()
last_verified_time = 0
# Configs
FACE_TIMEOUT = 60 # seconds of face absence before cancelling
VERIFY_INTERVAL = 30 # seconds between face identity checks
def is_confident_pose(face_roi):
"""Very basic heuristic — head is upright, eyes are visible."""
gray = cv2.cvtColor(face_roi, cv2.COLOR_BGR2GRAY)
eyes = eye_cascade.detectMultiScale(gray, 1.1, 4)
return len(eyes) >= 1
def match_identity(live_face, ref_img):
try:
result = DeepFace.verify(live_face, ref_img, enforce_detection=False, model_name='Facenet', detector_backend='opencv')
return result["verified"]
except Exception as e:
print("Verification error:", e)
return False
class VideoProcessor:
def __init__(self):
self.last_check = time.time()
self.last_verified = time.time()
self.face_missing = False
def recv(self, frame):
global face_detected_time, last_verified_time
img = frame.to_ndarray(format="bgr24")
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.1, 4)
if len(faces) == 0:
if time.time() - face_detected_time > FACE_TIMEOUT:
cv2.putText(img, "❌ Interview Cancelled: Face not visible!", (30, 30),
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 0, 255), 2)
else:
cv2.putText(img, "⚠️ Face not visible. You have 60 seconds.", (30, 30),
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 255), 2)
else:
face_detected_time = time.time()
for (x, y, w, h) in faces:
face_roi = img[y:y+h, x:x+w]
# Eye and posture check
confident = is_confident_pose(face_roi)
status = "✅ Confident Pose" if confident else "⚠️ Look Straight!"
color = (0, 255, 0) if confident else (0, 255, 255)
cv2.putText(img, status, (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, color, 2)
cv2.rectangle(img, (x, y), (x + w, y + h), color, 2)
# Identity check every VERIFY_INTERVAL seconds
if uploaded_image and (time.time() - last_verified_time) > VERIFY_INTERVAL:
matched = match_identity(face_roi, user_ref_img)
last_verified_time = time.time()
if not matched:
cv2.putText(img, "❌ Identity mismatch!", (x, y + h + 30),
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 0, 255), 2)
else:
cv2.putText(img, "✅ Identity Verified", (x, y + h + 30),
cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 0), 2)
return av.VideoFrame.from_ndarray(img, format="bgr24")
webrtc_streamer(
key="monitor",
video_processor_factory=VideoProcessor,
mode=WebRtcMode.RECVONLY, # ✅ FIXED: Use enum, not string
rtc_configuration=RTCConfiguration(iceServers=[]) # Disable STUN
) |