server.py
7.53 KB
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
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
import os
import torch
import numpy as np
import asyncio
import json
import base64
import websockets
from io import BytesIO
import pymysql
from datetime import datetime
from PIL import Image, ImageDraw
from IPython import display
from models.mtcnn import MTCNN
from models.inception_resnet_v1 import InceptionResnetV1
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
print('Running on device: {}'.format(device))
model = InceptionResnetV1().eval().to(device)
attendance_db = pymysql.connect(
user='root',
passwd='',
host='localhost',
db='attendance',
charset='utf8'
)
lock = asyncio.Lock()
clients = set()
#processes = []
async def get_embeddings(face_list):
global model
x = torch.Tensor(face_list).to(device)
yhat = model(x)
return yhat
async def get_distance(arr1, arr2):
distance = np.linalg.norm(arr1 - arr2)
return distance
async def get_cosine_similarity(arr1, arr2):
similarity = np.inner(arr1, arr2) / (np.linalg.norm(arr1) * np.linalg.norm(arr2))
return similarity
async def register(websocket):
global lock
global clients
async with lock:
clients.add(websocket)
remote_ip = websocket.remote_address[0]
msg='[{ip}] connected'.format(ip=remote_ip)
print(msg)
async def unregister(websocket):
global lock
global clients
async with lock:
clients.remove(websocket)
remote_ip = websocket.remote_address[0]
msg='[{ip}] disconnected'.format(ip=remote_ip)
print(msg)
async def thread(websocket, path):
await register(websocket)
try:
async for message in websocket:
data = json.loads(message)
remote_ip = websocket.remote_address[0]
if data['action'] == 'register':
# log
msg='[{ip}] register face'.format(ip=remote_ip)
print(msg)
# load json
student_id = data['student_id']
student_name = data['student_name']
face = np.asarray(data['MTCNN'], dtype = np.float32)
face = face.reshape((1,3,160,160))
# DB에 연결
cursor = attendance_db.cursor(pymysql.cursors.DictCursor)
# 학생을 찾음
sql = "SELECT student_id FROM student WHERE student_id = %s;"
cursor.execute(sql, (student_id))
# DB에 학생이 없으면 등록
if not cursor.fetchone():
sql = "INSERT INTO student(student_id, student_name) VALUES (%s, %s)"
cursor.execute(sql, (student_id, student_name))
attendance_db.commit()
msg='[{ip}] {id} is registered'.format(ip=remote_ip, id=student_id)
print(msg)
# student_embedding Table에 등록
embedding = await get_embeddings(face)
embedding = embedding.detach().numpy().tobytes()
embedding_date = datetime.now().strftime('%Y-%m-%d')
sql = "insert into student_embedding(student_id, embedding_date, embedding) values (%s, %s, _binary %s)"
cursor.execute(sql, (student_id, embedding_date, embedding))
attendance_db.commit()
send = json.dumps({'status': 'success', 'student_id': student_id})
await websocket.send(send)
elif data['action'] == 'verify':
# log
msg='[{ip}] verify face'.format(ip=remote_ip)
print(msg)
# load json
face = np.asarray(data['MTCNN'], dtype = np.float32)
face = face.reshape((1,3,160,160))
embedding = await get_embeddings(face)
embedding = embedding.detach().numpy()
# 가장 비슷한 Embedding을 찾는 SQL
cursor = attendance_db.cursor(pymysql.cursors.DictCursor)
sql = "SELECT student_id, embedding FROM student_embedding;"
cursor.execute(sql)
result = cursor.fetchall()
verified_id = '0'
distance_min = 99
for row_data in result:
db_embedding = np.frombuffer(row_data['embedding'], dtype=np.float32)
db_embedding = db_embedding.reshape((1,512))
distance = await get_distance(embedding, db_embedding)
if (distance < distance_min):
verified_id = row_data['student_id']
distance_min = distance
# 출석 데이터 전송
print('[debug] distance:', distance_min)
send = ''
if distance_min < 0.4:
# 인증 성공
# 오늘 이미 출석 됐는지 확인
sql = "SELECT DATE(timestamp) FROM student_attendance WHERE (lecture_id=%s) AND (student_id=%s) AND (DATE(timestamp) = CURDATE());"
cursor.execute(sql, ('0', verified_id))
# 출석 기록이 없는 경우에만
if not cursor.fetchone():
# 테이블 맨 뒤에 datetime attribute가 있음. 서버 시간 가져오게 default로 설정해둠.
sql = "INSERT INTO student_attendance(lecture_id, student_id, status) VALUES (%s, %s, %s)"
# TODO: attend / late 처리
cursor.execute(sql, ('0', verified_id, 'attend'))
attendance_db.commit()
# log 작성
msg='[{ip}] verification success {id}'.format(ip=remote_ip, id=verified_id)
print(msg)
send = json.dumps({'status': 'success', 'student_id': verified_id})
else:
msg='[{ip}] verification failed: {id} is already verified'.format(ip=remote_ip, id=verified_id)
print(msg)
send = json.dumps({'status': 'already', 'student_id': verified_id})
else:
# 인증 실패
msg='[{ip}] verification failed'.format(ip=remote_ip)
print(msg)
send = json.dumps({'status': 'fail'})
await websocket.send(send)
elif data['action'] == "save_image":
# 출석이 제대로 이뤄지지 않으면 이미지를 저장하여
# 나중에 교강사가 출석을 확인할 수 있도록 한다
msg='[{ip}] save image'.format(ip=remote_ip)
print(msg)
arr = np.asarray(data['image'], dtype = np.uint8)
blob = arr.tobytes()
# TODO: lecture DB에 tuple 삽입해야 아래 코드가 돌아감
# 테이블 맨 뒤에 datetime attribute가 있음. 서버 시간 가져오게 default로 설정해둠.
cursor = attendance_db.cursor(pymysql.cursors.DictCursor)
sql = "INSERT INTO undefined_image(lecture_id, image, width, height) VALUES (%s, _binary %s, %s, %s)"
cursor.execute(sql, ('0', blob, arr.shape[0], arr.shape[1]))
attendance_db.commit()
else:
print("unsupported event: {}", data)
finally:
await unregister(websocket)
print('run verification server')
start_server = websockets.serve(thread, '0.0.0.0', 8765)
asyncio.get_event_loop().run_until_complete(start_server)
asyncio.get_event_loop().run_forever()