import tensorflow as tf
import cv2
import numpy as np
from glob import glob
# from models import Yolov4
import gradio as gr
# model = Yolov4(weight_path="best.pt", class_name_path='coco_classes.txt')


# from ultralytics import YOLO

# Load a model
# model = YOLO("best.pt")  # load a custom model

# Predict with the model
# results = model("image.jpg", save = True)  # predict on an image

face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')

def detect_faces(frame):
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    faces = face_cascade.detectMultiScale(gray, 1.3, 5)
    print(f"Detected {len(faces)} faces")
    return len(faces)

def detect_faces_in_video():
    success, frame = camera.read()
    if success:
        num_faces = detect_faces(frame)
        return int(num_faces)
    else:
        return None

"""def gradio_wrapper(img):
    global face_cascade
    #print(np.shape(img))
    results = model.predict(img)  # predict on an image
    try:
        if max(results[0].boxes.cls) == 0:
            text = "Man"
        if max(results[0].boxes.cls) == 1:
            text = "Women"
    except:
        pass
        
    return cv2.putText(img, text,(00, 185), cv2.FONT_HERSHEY_SIMPLEX, 1, 
                  (0, 0, 255),  2, cv2.LINE_AA, False)
    # return results

"""
  
demo = gr.Interface(
    detect_faces_in_video,
    #gr.Image(source="webcam", streaming=True, flip=True),
    gr.Image(source="webcam", streaming=True),
    "image",
    live=True
)

demo.launch()