-
Notifications
You must be signed in to change notification settings - Fork 0
/
detect.py
224 lines (186 loc) · 7.99 KB
/
detect.py
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
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
import csv
import pandas as pd
import argparse
import generate_vector
from sys import platform
from models import * # set ONNX_EXPORT in models.py
from utils.datasets import *
from utils.utils import *
import cv2
import numpy as np
import sys
import subprocess
import time
import os
from os import walk
from os import listdir
from os.path import isfile, join
import glob
def initialize(names):
ad_path = os.getcwd()+os.sep+"ads"
onlyfiles = [f for f in listdir(ad_path) if isfile(join(ad_path, f))]
ad_url = [x for x in onlyfiles if x.endswith('mp4')]
length = len(ad_path+os.sep)
for i in range(len(ad_url)):
ad_url[i] = ad_path+os.sep+ad_url[i]
if not (os.path.exists("signature.csv")):
opt.status = "delete"
if(opt.status == "delete"):
if(os.path.exists("signature.csv")):
os.remove("signature.csv")
names = names.tolist()
names.insert(0,"Ad Filename")
names.insert(1,"Ad Frames")
names.insert(2, "persons")
with open("signature.csv","a",newline='') as my_csv:
csvWriter = csv.writer(my_csv,delimiter=',')
if (opt.status == "delete"):
csvWriter.writerow(names)
for ad in ad_url:
cap = cv2.VideoCapture(ad)
length_of_ad = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))-125
print("\nInitializing..."+str(ad[length:]))
csvWriter.writerow([ad[length:]]+[length_of_ad]+generate_vector.detect(ad, 608, 5))
def rmse(arr, desc_ad):
x = 0
for i in range(len(arr)):
x += pow((arr[i] - desc_ad[i]), 2)
x = pow(x, 0.5)
return x
def detect(desc, ad_name, ad_length):
img_size = 608
source, weights, half = opt.source, 'weights/coco.pt', False
webcam = source == '0' or source.startswith('rtsp') or source.startswith('http') or source.endswith('.txt')
# Initialize
device = torch_utils.select_device(device='0')
# Initialize model
model = Darknet('cfg/yolov3-spp.cfg', img_size)
# Load weights
model.load_state_dict(torch.load(weights, map_location=device)['model'])
# Second-stage classifier
classify = False
if classify:
modelc = torch_utils.load_classifier(name='resnet101', n=2) # initialize
modelc.load_state_dict(torch.load('weights/resnet101.pt', map_location=device)['model']) # load weights
modelc.to(device).eval()
# Eval mode
model.to(device).eval()
# Export mode
if ONNX_EXPORT:
model.fuse()
img = torch.zeros((1, 3) + img_size) # (1, 3, 320, 192)
torch.onnx.export(model, img, 'weights/export.onnx', verbose=False, opset_version=11)
# Validate exported model
import onnx
model = onnx.load('weights/export.onnx') # Load the ONNX model
onnx.checker.check_model(model) # Check that the IR is well formed
print(onnx.helper.printable_graph(model.graph)) # Print a human readable representation of the graph
return
# Half precision
half = half and device.type != 'cpu' # half precision only supported on CUDA
if half:
model.half()
dataset = LoadImages(source, img_size=img_size, half=half)
# Get names
names = load_classes('data/coco.names')
# Run inference
sliding_box = []
arr = []
arr_temp = [0]*80
frame_num = 0
skip_duration = 0
t0 = time.time()
for path, img, im0s, vid_cap in dataset:
if(skip_duration>0):
skip_duration -= 1
frame_num += 1
continue
else:
t1_start = time.time()
# Get detections
img = torch.from_numpy(img).to(device)
if img.ndimension() == 3:
img = img.unsqueeze(0)
pred = model(img)[0]
if half:
pred = pred.float()
# Apply NMS
pred = non_max_suppression(pred, 0.3, 0.5, classes=None, agnostic=False)
# Apply Classifier
if classify:
pred = apply_classifier(pred, modelc, img, im0s)
# Process detections
for i, det in enumerate(pred): # detections per image
p, s, im0 = path, '', im0s
ih, iw = im0.shape[:2]
iht = int((0.15 * ih))
ihb = int(ih - (0.15*ih))
iwl = int((0.15 * iw))
iwr = int(iw - (0.15 * iw))
im0 = im0[iht:ihb, iwl:iwr]
frame_num += 1
# s += '%gx%g ' % img.shape[2:] # print string
# print(ad_name)
if det is not None and len(det):
# Rescale boxes from img_size to im0 size
det[:, :4] = scale_coords(img.shape[2:], det[:, :4], im0.shape).round()
# Print results
for c in det[:, -1].unique():
n = (det[:, -1] == c).sum() # detections per class
arr_temp[int(c)] += int(n)
# s += '%g %ss, ' % (n, names[int(c)]) # add to string
t1_end = time.time()
t2_start = time.time()
arr.append(arr_temp)
if(frame_num<125):
sliding_box = np.sum(arr, axis = 0)
elif(frame_num==125):
sliding_box = np.sum(arr, axis = 0)
for i in range(len(desc)):
if((sliding_box==desc[i]).all()):
skip_duration = ad_length[i]
print(ad_name[i], end=", ")
print(time.strftime("%H:%M:%S", time.gmtime((frame_num - 125)/25)-25)+"\n")
break
else:
sliding_box = np.subtract(sliding_box, arr[0])
sliding_box = np.add(sliding_box, arr[len(arr)-1])
del arr[0]
for i in range(len(desc)):
if((sliding_box==desc[i]).all()):
skip_duration = ad_length[i]
name = ad_name[i]
print("\n\n"+name, end=" --> ")
timestamp = time.strftime("%H:%M:%S", time.gmtime((frame_num - 125)/25))
print(timestamp, end=" --> ")
with open("result.txt", "a+") as f:
f.write(name+" --> "+str(timestamp)+"\n\n")
print("Skipping "+str(ad_length[i])+" frames\n\n")
break
# Printitng the time required to process each frame, for real time, it should be bwlow 0.04
print((t1_end - t1_start),(time.time() - t2_start), end = '\n')
arr_temp = [0]*80
print('Done. (%.3fs)' % (time.time() - t0))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--source', type=str, default='footage.mp4', help='source') # input file/folder, 0 for webcam
parser.add_argument('--status', type=str, default="keep", help='keep, append, delete')
opt = parser.parse_args()
# print(opt)
df = pd.read_csv("coco.csv")
names = df.values
names = names.reshape((1,79)).flatten()
if(os.path.exists("result.txt")):
os.remove("result.txt")
# if(os.path.exists("rmse.txt")):
# os.remove("rmse.txt")
with torch.no_grad():
if(opt.status == "append" or opt.status == "delete" or not(os.path.exists("signature.csv"))):
initialize(names)
df = pd.read_csv("signature.csv")
ad_name = df.iloc[:,0].values
ad_length = df.iloc[:,1].values
desc = df.iloc[:,2:].values
print(ad_name)
print("\nInitialization Complete....Now starting detection\n")
detect(desc, ad_name, ad_length)