restime = reciprocal_of_max_frame_rate - loop.time() + start_time
if restime > 0:
await asyncio.sleep(restime)
# =================== INIT ====================
# address_dict = ['10.41.0.198', '10.41.0.199']
address_dict = ['10.41.0.231']
os.environ['MXNET_CUDNN_AUTOTUNE_DEFAULT'] = '0'
frame_buffer_size = 10 * len(address_dict)
upstream_frame_queue = Queue(maxsize=frame_buffer_size)
suspicion_face_queue = Queue(maxsize=frame_buffer_size)
result_queue = Queue(maxsize=frame_buffer_size)
# =================== ARGS ====================
args = start_up_init()
args.address_dict = address_dict
# =================== Process On ====================
args.ip_address = '10.41.0.231'
frame_queue_231 = Queue(maxsize=frame_buffer_size)
Process(target=lambda: asyncio.run(
detection_loop(args, frame_queue_231))).start()
# args.ip_address = '10.41.0.232'
# frame_queue_232 = Queue(maxsize=frame_buffer_size)
# Process(target=lambda: asyncio.run(
# detection_loop(args, frame_queue_232))).start()
Process(target=lambda: asyncio.run(embedding_loop(args))).start()
Process(target=lambda: asyncio.run(camera_loop(args))).start()
start_time = loop.time()
camera.read()[1]
# pairs = [(ip, camera_dict[ip].frame(rows=672, cols=672)) for ip in addr_list]
# pairs = [(str(code), camera_dict[code].read()[1]) for code in code_list]
pairs = [(str(code), cv2.resize(camera_dict[code].read()[1],
(672, 672))) for code in code_list]
frame_queue.put(pairs)
#print(loop.time() - start_time, frame_queue.qsize())
restime = rate - loop.time() + start_time
if restime > 0:
await asyncio.sleep(restime)
# =================== INIT ====================
os.environ["MXNET_CUDNN_AUTOTUNE_DEFAULT"] = "0"
preload = start_up_init()
# preload.scales = [0.7]
# preload.usb_camera_code = [0]
preload.max_face_number = 30
print(preload.usb_camera_code)
camera_dict = {}
for code in preload.usb_camera_code:
camera = cv2.VideoCapture(code)
# camera = cv2.VideoCapture('./trump.mp4') £ use like this for videos testing
camera_dict[code] = camera
# =================== QUEUE ====================
frame_buffer_size = preload.queue_buffer_size
frame_queue = Queue(frame_buffer_size)
import sys import numpy as np import time from termcolor import colored import matplotlib.pyplot as plt import pickle from sklearn.neural_network import MLPClassifier from sklearn.preprocessing import label_binarize from sklearn.model_selection import train_test_split from sklearn import metrics from helper import read_pkl_model, start_up_init, get_dataset, get_image_paths_and_labels # =================== ARGS ==================== args = start_up_init(True) # =================== MODEL CLASS ==================== arcface = face_model.FaceModel(args) # =================== LOAD DATASET ====================. dir_train = './Embedding/train.npy' data_train = './Temp/train_data' dataset_train = get_dataset(data_train) paths_train, labels_train = get_image_paths_and_labels(dataset_train) dir_safe = './Embedding/safe.npy' data_safe = './Temp/safe_data' dataset_safe = get_dataset(data_safe) paths_safe, labels_safe = get_image_paths_and_labels(dataset_safe)