class GN():
def __init__(self, seq_index, tt, cuda=True):
'''
Evaluating with the MotMetrics
:param seq_index: the number of the sequence
:param tt: train_test
:param length: the number of frames which is used for training
:param cuda: True - GPU, False - CPU
'''
self.bbx_counter = 0
self.seq_index = seq_index
self.hungarian = Munkres()
self.device = torch.device("cuda" if cuda else "cpu")
self.tt = tt
self.missingCounter = 0
self.sideConnection = 0
print ' Loading the model...'
self.loadModel()
self.out_dir = t_dir + 'motmetrics_%s/' % type
if not os.path.exists(self.out_dir):
os.mkdir(self.out_dir)
else:
deleteDir(self.out_dir)
os.mkdir(self.out_dir)
self.initOut()
def initOut(self):
print ' Loading Data...'
self.train_set = DatasetFromFolder(
sequence_dir, '../MOT/MOT16/test/MOT16-%02d' % self.seq_index)
detection_dir = self.out_dir + 'res_training_det.txt'
res_training = self.out_dir + 'res_training.txt' # the result of the training data
self.createTxt(detection_dir)
self.createTxt(res_training)
self.copyLines(self.seq_index, 1, detection_dir, self.tt, 1)
self.evaluation(1, self.tt, detection_dir, res_training)
def getSeqL(self, info):
# get the length of the sequence
f = open(info, 'r')
f.readline()
for line in f.readlines():
line = line.strip().split('=')
if line[0] == 'seqLength':
seqL = int(line[1])
f.close()
return seqL
def copyLines(self, seq, head, gt_seq, tail=-1, tag=0):
'''
Copy the groun truth within [head, head+num]
:param seq: the number of the sequence
:param head: the head frame number
:param tail: the number the clipped sequence
:param gt_seq: the dir of the output file
:return: None
'''
if tt_tag:
basic_dir = '../MOT/MOT%d/test/MOT%d-%02d-%s/' % (year, year, seq,
type)
else:
basic_dir = '../MOT/MOT%d/train/MOT%d-%02d-%s/' % (year, year, seq,
type)
print ' Testing on', basic_dir, 'Length:', self.tt
seqL = tail if tail != -1 else self.getSeqL(basic_dir + 'seqinfo.ini')
det_dir = 'gt/gt_det.txt' if test_gt_det else 'det/det.txt'
seq_dir = basic_dir + ('gt/gt.txt' if tag == 0 else det_dir)
inStream = open(seq_dir, 'r')
outStream = open(gt_seq, 'w')
for line in inStream.readlines():
line = line.strip()
attrs = line.split(',')
f_num = int(attrs[0])
if f_num >= head and f_num <= seqL:
print >> outStream, line
outStream.close()
inStream.close()
return seqL
def createTxt(self, out_file):
f = open(out_file, 'w')
f.close()
def loadModel(self):
name = 'all_7'
if edge_initial == 1:
i_name = 'Random/'
elif edge_initial == 0:
i_name = 'IoU/'
self.Uphi = torch.load('Results/MOT16/%s/%s/uphi_13.pth' %
(i_name, name)).to(self.device)
self.Ephi = torch.load('Results/MOT16/%s/%s/ephi_13.pth' %
(i_name, name)).to(self.device)
self.u = torch.load('Results/MOT16/%s/%s/u_13.pth' % (i_name, name))
self.u = self.u.to(self.device)
def swapFC(self):
self.cur = self.cur ^ self.nxt
self.nxt = self.cur ^ self.nxt
self.cur = self.cur ^ self.nxt
def linearModel(self, out, attr1, attr2):
# print 'I got you! *.*'
t = attr1[-1]
self.sideConnection += 1
if t > f_gap:
return
frame = int(attr1[0])
x1, y1, w1, h1 = float(attr1[2]), float(attr1[3]), float(
attr1[4]), float(attr1[5])
x2, y2, w2, h2 = float(attr2[2]), float(attr2[3]), float(
attr2[4]), float(attr2[5])
x_delta = (x2 - x1) / t
y_delta = (y2 - y1) / t
w_delta = (w2 - w1) / t
h_delta = (h2 - h1) / 2
for i in xrange(1, t):
frame += 1
x1 += x_delta
y1 += y_delta
w1 += w_delta
h1 += h_delta
attr1[0] = str(frame)
attr1[2] = str(x1)
attr1[3] = str(y1)
attr1[4] = str(w1)
attr1[5] = str(h1)
line = ''
for attr in attr1[:-1]:
line += attr + ','
if show_recovering:
line += '1'
else:
line = line[:-1]
print >> out, line
self.bbx_counter += 1
self.missingCounter += t - 1
def evaluation(self, head, tail, gtFile, outFile):
'''
Evaluation on dets
:param head: the head frame number
:param tail: the tail frame number
:param gtFile: the ground truth file name
:param outFile: the name of output file
:return: None
'''
gtIn = open(gtFile, 'r')
self.cur, self.nxt = 0, 1
line_con = [[], []]
id_con = [[], []]
id_step = 1
step = head + self.train_set.setBuffer(head)
while step < tail:
# print '*********************************'
t_gap = self.train_set.loadNext()
step += t_gap
# print head+step, 'F',
u_ = self.Uphi(self.train_set.E, self.train_set.V, self.u)
# print 'Fo'
m = self.train_set.m
n = self.train_set.n
# print 'm = %d, n = %d'%(m, n)
if n == 0:
print 'There is no detection in the rest of sequence!'
break
if id_step == 1:
out = open(outFile, 'a')
i = 0
while i < m:
attrs = gtIn.readline().strip().split(',')
if float(attrs[6]) >= tau_conf_score:
attrs.append(1)
attrs[1] = str(id_step)
line = ''
for attr in attrs[:-1]:
line += attr + ','
if show_recovering:
line += '0'
else:
line = line[:-1]
print >> out, line
self.bbx_counter += 1
line_con[self.cur].append(attrs)
id_con[self.cur].append(id_step)
id_step += 1
i += 1
out.close()
i = 0
while i < n:
attrs = gtIn.readline().strip().split(',')
if float(attrs[6]) >= tau_conf_score:
attrs.append(1)
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(-1)
i += 1
# update the edges
# print 'T',
ret = self.train_set.getRet()
for edge in self.train_set.candidates:
e, vs_index, vr_index = edge
if ret[vs_index][vr_index] == 1.0:
continue
e = e.to(self.device).view(1, -1)
v1 = self.train_set.getMApp(1, vs_index).to(self.device)
v2 = self.train_set.getMApp(
0, vr_index, vs_index,
line_con[self.cur][vs_index][-1]).to(self.device)
e_ = self.Ephi(e, v1, v2, u_)
self.train_set.edges[vs_index][vr_index] = e_.data.view(-1)
tmp = F.softmax(e_)
tmp = tmp.cpu().data.numpy()[0]
ret[vs_index][vr_index] = float(tmp[0])
# self.train_set.showE(outFile)
# for j in ret:
# print j
results = self.hungarian.compute(ret)
out = open(outFile, 'a')
look_up = set(j for j in xrange(n))
for (i, j) in results:
# print (i,j)
if ret[i][j] >= tau_threshold:
continue
look_up.remove(j)
self.train_set.updateVelocity(i, j, line_con[self.cur][i][-1],
False)
id = id_con[self.cur][i]
id_con[self.nxt][j] = id
attr1 = line_con[self.cur][i]
attr2 = line_con[self.nxt][j]
# print attrs
attr2[1] = str(id)
if attr1[-1] > 1:
# for the missing detections
self.linearModel(out, attr1, attr2)
line = ''
for attr in attr2[:-1]:
line += attr + ','
if show_recovering:
line += '0'
else:
line = line[:-1]
print >> out, line
self.bbx_counter += 1
for j in look_up:
self.train_set.updateVelocity(-1, j, tag=False)
for i in xrange(n):
if id_con[self.nxt][i] == -1:
id_con[self.nxt][i] = id_step
attrs = line_con[self.nxt][i]
attrs[1] = str(id_step)
line = ''
for attr in attrs[:-1]:
line += attr + ','
if show_recovering:
line += '0'
else:
line = line[:-1]
print >> out, line
self.bbx_counter += 1
id_step += 1
out.close()
# For missing & Occlusion
index = 0
for (i, j) in results:
while i != index:
attrs = line_con[self.cur][index]
# print '*', attrs, '*'
if attrs[-1] + t_gap <= gap:
attrs[-1] += t_gap
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(id_con[self.cur][index])
self.train_set.moveMApp(index)
index += 1
if ret[i][j] >= tau_threshold:
attrs = line_con[self.cur][index]
# print '*', attrs, '*'
if attrs[-1] + t_gap <= gap:
attrs[-1] += t_gap
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(id_con[self.cur][index])
self.train_set.moveMApp(index)
index += 1
while index < m:
attrs = line_con[self.cur][index]
# print '*', attrs, '*'
if attrs[-1] + t_gap <= gap:
attrs[-1] += t_gap
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(id_con[self.cur][index])
self.train_set.moveMApp(index)
index += 1
# con = self.train_set.cleanEdge()
# for i in xrange(len(con)-1, -1, -1):
# index = con[i]
# del line_con[self.nxt][index]
# del id_con[self.nxt][index]
line_con[self.cur] = []
id_con[self.cur] = []
# print head+step, results
self.train_set.swapFC()
self.swapFC()
gtIn.close()
print ' The results:', id_step, self.bbx_counter
class GN():
def __init__(self, seq_index, tt, length, cuda=True):
'''
Evaluating with the MotMetrics
:param seq_index: the number of the sequence
:param tt: train_test
:param length: the number of frames which is used for training
:param cuda: True - GPU, False - CPU
'''
self.seq_index = seq_index
self.hungarian = Munkres()
self.device = torch.device("cuda" if cuda else "cpu")
self.tt = tt
self.length = length
self.missingCounter = 0
self.sideConnection = 0
print ' Loading the model...'
self.loadModel()
self.out_dir = t_dir + 'motmetrics/'
if not os.path.exists(self.out_dir):
os.mkdir(self.out_dir)
else:
deleteDir(self.out_dir)
os.mkdir(self.out_dir)
self.initOut()
def initOut(self):
start = time.time()
print ' Loading Data...'
print ' Training'
self.train_set = DatasetFromFolder(sequence_dir)
gt_training = self.out_dir + 'gt_training.txt' # the gt of the training data
self.copyLines(self.seq_index, 1, gt_training, self.tt)
detection_dir = self.out_dir + 'res_training_det.txt'
res_training = self.out_dir + 'res_training.txt' # the result of the training data
self.createTxt(detection_dir)
self.createTxt(res_training)
self.copyLines(self.seq_index, 1, detection_dir, self.tt, 1)
# Evaluating on the training data
# motmetrics = open(metrics_dir, 'a')
# print >> motmetrics, '*'*30, self.tt, '*'*30
# print >> motmetrics, 'Training'
self.evaluation(1, self.tt, detection_dir, res_training)
print ' Time consuming:', (time.time() - start) / 60.0
# cmd = 'python3 evaluation.py %s %s'%(gt_training, res_training)
# (status, output) = commands.getstatusoutput(cmd)
# print >> motmetrics, output
# print >> motmetrics, 'The time consuming:{}\n\n'.format((time.time()-start)/60)
# motmetrics.close()
if self.tt < self.length:
# Evaluating on the validation data
start = time.time()
print ' Validation'
# The distant sequence
head = self.length - self.tt + 1
tail = self.length
# The sequence nearby
# head = self.tt
# tail = 2*self.tt-1
gt_valiadation = self.out_dir + 'gt_validation.txt' # the gt of the validation data
self.copyLines(self.seq_index, head, gt_valiadation, tail)
detection_dir = self.out_dir + 'res_validation_det.txt'
res_validation = self.out_dir + 'res_validation.txt' # the result of the validation data
self.createTxt(detection_dir)
self.createTxt(res_validation)
self.copyLines(self.seq_index, head, detection_dir, tail, 1)
# motmetrics = open(metrics_dir, 'a')
# print >> motmetrics, 'Validation'
self.evaluation(head, tail, detection_dir, res_validation)
print ' Time consuming:', (time.time() - start) / 60.0
# cmd = 'python3 evaluation.py %s %s'%(gt_valiadation, res_validation)
# (status, output) = commands.getstatusoutput(cmd)
# print >> motmetrics, output
# print >> motmetrics, 'The time consuming:{}\n\n'.format((time.time()-start)/60)
# motmetrics.close()
else:
# Evaluating on the validation data
for seq in seqs:
if seq == self.seq_index:
continue
print ' %02d_Validation' % seq
start = time.time()
seq_dir = 'MOT16/train/MOT%d-%02d' % (year, seq)
self.train_set = DatasetFromFolder(seq_dir)
gt_seq = self.out_dir + 'gt_%02d.txt' % seq
seqL = self.copyLines(seq, 1, gt_seq)
detection_dir = self.out_dir + 'res_%02d_det.txt' % seq
c_validation = self.out_dir + 'res_%02d.txt' % seq
self.createTxt(detection_dir)
self.createTxt(c_validation)
self.copyLines(seq, 1, detection_dir, tag=1)
# motmetrics = open(metrics_dir, 'a')
# print >> motmetrics, '%02d_validation'%seq
self.evaluation(1, seqL, detection_dir, c_validation)
print ' Time consuming:', (time.time() - start) / 60.0
# cmd = 'python3 evaluation.py %s %s'%(gt_seq, c_validation)
# (status, output) = commands.getstatusoutput(cmd)
# print >> motmetrics, output
# print >> motmetrics, 'The time consuming:{}\n\n'.format((time.time()-start)/60)
# motmetrics.close()
def getSeqL(self, info):
# get the length of the sequence
f = open(info, 'r')
f.readline()
for line in f.readlines():
line = line.strip().split('=')
if line[0] == 'seqLength':
seqL = int(line[1])
f.close()
return seqL
def copyLines(self, seq, head, gt_seq, tail=-1, tag=0):
'''
Copy the groun truth within [head, head+num]
:param seq: the number of the sequence
:param head: the head frame number
:param tail: the number the clipped sequence
:param gt_seq: the dir of the output file
:return: None
'''
basic_dir = 'MOT16/train/MOT%d-%02d/' % (year, seq)
seqL = tail if tail != -1 else self.getSeqL(basic_dir + 'seqinfo.ini')
det_dir = 'gt/gt_det.txt' if test_gt_det else 'det/det.txt'
seq_dir = basic_dir + ('gt/gt.txt' if tag == 0 else det_dir)
inStream = open(seq_dir, 'r')
outStream = open(gt_seq, 'w')
for line in inStream.readlines():
line = line.strip()
attrs = line.split(',')
f_num = int(attrs[0])
if f_num >= head and f_num <= seqL:
print >> outStream, line
outStream.close()
inStream.close()
return seqL
def createTxt(self, out_file):
f = open(out_file, 'w')
f.close()
def loadModel(self):
# self.Uphi = torch.load(t_dir+'uphi.pth').to(self.device)
# self.Ephi = torch.load(t_dir+'ephi.pth').to(self.device)
# self.u = torch.load(t_dir+'u.pth')
# self.u = self.u.to(self.device)
self.Uphi = torch.load('Results/MOT16/IoU/all/uphi_13.pth').to(
self.device)
self.Ephi = torch.load('Results/MOT16/IoU/all/ephi_13.pth').to(
self.device)
self.u = torch.load('Results/MOT16/IoU/all/u_13.pth')
self.u = self.u.to(self.device)
def swapFC(self):
self.cur = self.cur ^ self.nxt
self.nxt = self.cur ^ self.nxt
self.cur = self.cur ^ self.nxt
def linearModel(self, out, attr1, attr2):
# print 'I got you! *.*'
t = attr1[-1]
self.sideConnection += 1
if t > f_gap:
return
frame = int(attr1[0])
x1, y1, w1, h1 = float(attr1[2]), float(attr1[3]), float(
attr1[4]), float(attr1[5])
x2, y2, w2, h2 = float(attr2[2]), float(attr2[3]), float(
attr2[4]), float(attr2[5])
x_delta = (x2 - x1) / t
y_delta = (y2 - y1) / t
w_delta = (w2 - w1) / t
h_delta = (h2 - h1) / t
for i in xrange(1, t):
frame += 1
x1 += x_delta
y1 += y_delta
w1 += w_delta
h1 += h_delta
attr1[0] = str(frame)
attr1[2] = str(x1)
attr1[3] = str(y1)
attr1[4] = str(w1)
attr1[5] = str(h1)
line = ''
for attr in attr1[:-1]:
line += attr + ','
line = line[:-1]
print >> out, line
self.missingCounter += t - 1
def evaluation(self, head, tail, gtFile, outFile):
'''
Evaluation on dets
:param head: the head frame number
:param tail: the tail frame number
:param gtFile: the ground truth file name
:param outFile: the name of output file
:return: None
'''
gtIn = open(gtFile, 'r')
self.cur, self.nxt = 0, 1
line_con = [[], []]
id_con = [[], []]
id_step = 1
step = head + self.train_set.setBuffer(head)
while step < tail:
t_gap = self.train_set.loadNext()
step += t_gap
# print head+step, 'F',
u_ = self.Uphi(self.train_set.E, self.train_set.V, self.u)
# print 'Fo'
m = self.train_set.m
n = self.train_set.n
if n == 0:
print 'There is no detection in the rest of sequence!'
break
if id_step == 1:
out = open(outFile, 'a')
i = 0
while i < m:
attrs = gtIn.readline().strip().split(',')
if float(attrs[6]) >= tau_conf_score:
attrs.append(1)
attrs[1] = str(id_step)
line = ''
for attr in attrs[:-1]:
line += attr + ','
line = line[:-1]
print >> out, line
line_con[self.cur].append(attrs)
id_con[self.cur].append(id_step)
id_step += 1
i += 1
out.close()
i = 0
while i < n:
attrs = gtIn.readline().strip().split(',')
if float(attrs[6]) >= tau_conf_score:
attrs.append(1)
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(-1)
i += 1
# update the edges
# print 'T',
ret = self.train_set.getRet()
for edge in self.train_set.candidates:
e, vs_index, vr_index = edge
if ret[vs_index][vr_index] == 1.0:
continue
e = e.to(self.device).view(1, -1)
v1 = self.train_set.getMotion(1, vs_index)
v2 = self.train_set.getMotion(0, vr_index, vs_index)
e_ = self.Ephi(e, v1, v2, u_)
self.train_set.edges[vs_index][vr_index] = e_.data.view(-1)
tmp = F.softmax(e_)
tmp = tmp.cpu().data.numpy()[0]
ret[vs_index][vr_index] = float(tmp[0])
# self.train_set.showE(outFile)
# for j in ret:
# print j
results = self.hungarian.compute(ret)
out = open(outFile, 'a')
for (i, j) in results:
# print (i,j)
if ret[i][j] >= tau_threshold:
continue
id = id_con[self.cur][i]
id_con[self.nxt][j] = id
attr1 = line_con[self.cur][i]
attr2 = line_con[self.nxt][j]
# print attrs
attr2[1] = str(id)
if attr1[-1] > 1:
# for the missing detections
self.linearModel(out, attr1, attr2)
line = ''
for attr in attr2[:-1]:
line += attr + ','
line = line[:-1]
print >> out, line
for i in xrange(n):
if id_con[self.nxt][i] == -1:
id_con[self.nxt][i] = id_step
attrs = line_con[self.nxt][i]
attrs[1] = str(id_step)
line = ''
for attr in attrs[:-1]:
line += attr + ','
line = line[:-1]
print >> out, line
id_step += 1
out.close()
self.train_set.getVelocity(results)
index = 0
for (i, j) in results:
while i != index:
attrs = line_con[self.cur][index]
# print '*', attrs, '*'
if attrs[-1] + t_gap <= gap:
attrs[-1] += t_gap
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(id_con[self.cur][index])
self.train_set.moveMotion(index)
index += 1
index += 1
while index < m:
attrs = line_con[self.cur][index]
# print '*', attrs, '*'
if attrs[-1] + t_gap <= gap:
attrs[-1] += t_gap
line_con[self.nxt].append(attrs)
id_con[self.nxt].append(id_con[self.cur][index])
self.train_set.moveMotion(index)
index += 1
line_con[self.cur] = []
id_con[self.cur] = []
# print head+step, results
self.train_set.swapFC()
self.swapFC()
gtIn.close()
class GN():
def __init__(self, seq_index, tt, length, cuda=True):
'''
Evaluating with the MotMetrics
:param seq_index: the number of the sequence
:param tt: train_test
:param length: the number of frames which is used for training
:param cuda: True - GPU, False - CPU
'''
# top index: 0 - correct matching, 1 - false matching, second index: 0 - min, 1 - max, 2 - total, 3 - counter
self.ctau = [[1.0, 0.0, 0.0, 0] for i in xrange(2)] # get the threshold for matching cost
self.seq_index = seq_index
self.hungarian = Munkres()
self.device = torch.device("cuda" if cuda else "cpu")
self.tt = tt
self.length = length
self.missingCounter = 0
print ' Loading the model...'
self.loadModel()
self.out_dir = t_dir + 'motmetrics/'
if not os.path.exists(self.out_dir):
os.mkdir(self.out_dir)
else:
deleteDir(self.out_dir)
os.mkdir(self.out_dir)
self.initOut()
def initOut(self):
start = time.time()
print ' Loading Data...'
print ' Evaluating'
self.train_set = DatasetFromFolder(sequence_dir)
gt_training = self.out_dir + 'gt_training.txt' # the gt of the training data
self.copyLines(self.seq_index, 1, gt_training, self.tt)
detection_dir = self.out_dir +'res_training_det.txt'
res_training = self.out_dir + 'res_training.txt' # the result of the training data
self.createTxt(detection_dir)
self.createTxt(res_training)
self.copyLines(self.seq_index, 1, detection_dir, self.tt, 1)
# Evaluating on the training data
# motmetrics = open(metrics_dir, 'a')
# print >> motmetrics, '*'*30, self.tt, '*'*30
# print >> motmetrics, 'Training'
self.evaluation(1, self.tt, detection_dir)
print ' Time consuming:', (time.time()-start)/60.0
# cmd = 'python3 evaluation.py %s %s'%(gt_training, res_training)
# (status, output) = commands.getstatusoutput(cmd)
# print >> motmetrics, output
# print >> motmetrics, 'The time consuming:{}\n\n'.format((time.time()-start)/60)
# motmetrics.close()
if self.tt < self.length:
# Evaluating on the validation data
start = time.time()
print ' Validation'
# The distant sequence
head = self.length - self.tt + 1
tail = self.length
# The sequence nearby
# head = self.tt
# tail = 2*self.tt-1
gt_valiadation = self.out_dir + 'gt_validation.txt' # the gt of the validation data
self.copyLines(self.seq_index, head, gt_valiadation, tail)
detection_dir = self.out_dir + 'res_validation_det.txt'
res_validation = self.out_dir + 'res_validation.txt' # the result of the validation data
self.createTxt(detection_dir)
self.createTxt(res_validation)
self.copyLines(self.seq_index, head, detection_dir, tail, 1)
# motmetrics = open(metrics_dir, 'a')
# print >> motmetrics, 'Validation'
self.evaluation(head, tail, detection_dir)
print ' Time consuming:', (time.time()-start)/60.0
# cmd = 'python3 evaluation.py %s %s'%(gt_valiadation, res_validation)
# (status, output) = commands.getstatusoutput(cmd)
# print >> motmetrics, output
# print >> motmetrics, 'The time consuming:{}\n\n'.format((time.time()-start)/60)
# motmetrics.close()
else:
# Evaluating on the validation data
for seq in seqs:
if seq == self.seq_index:
continue
print ' %02d_Validation'%seq
start = time.time()
seq_dir = '../MOT/MOT16/train/MOT%d-%02d' % (year, seq)
self.train_set = DatasetFromFolder(seq_dir)
gt_seq = self.out_dir + 'gt_%02d.txt' % seq
seqL = self.copyLines(seq, 1, gt_seq)
detection_dir = self.out_dir + 'res_%02d_det.txt' % seq
c_validation = self.out_dir + 'res_%02d.txt' % seq
self.createTxt(detection_dir)
self.createTxt(c_validation)
self.copyLines(seq, 1, detection_dir, tag=1)
# motmetrics = open(metrics_dir, 'a')
# print >> motmetrics, '%02d_validation'%seq
self.evaluation(1, seqL, detection_dir)
print ' Time consuming:', (time.time()-start)/60.0
# cmd = 'python3 evaluation.py %s %s'%(gt_seq, c_validation)
# (status, output) = commands.getstatusoutput(cmd)
# print >> motmetrics, output
# print >> motmetrics, 'The time consuming:{}\n\n'.format((time.time()-start)/60)
# motmetrics.close()
def getSeqL(self, info):
# get the length of the sequence
f = open(info, 'r')
f.readline()
for line in f.readlines():
line = line.strip().split('=')
if line[0] == 'seqLength':
seqL = int(line[1])
f.close()
return seqL
def copyLines(self, seq, head, gt_seq, tail=-1, tag=0):
'''
Copy the groun truth within [head, head+num]
:param seq: the number of the sequence
:param head: the head frame number
:param tail: the number the clipped sequence
:param gt_seq: the dir of the output file
:return: None
'''
basic_dir = '../MOT/MOT16/train/MOT%d-%02d/' % (year, seq)
seqL = tail if tail != -1 else self.getSeqL(basic_dir + 'seqinfo.ini')
det_dir = 'gt/gt_det.txt' if test_gt_det else 'det/det.txt'
seq_dir = basic_dir + ('gt/gt.txt' if tag == 0 else det_dir)
inStream = open(seq_dir, 'r')
outStream = open(gt_seq, 'w')
for line in inStream.readlines():
line = line.strip()
attrs = line.split(',')
f_num = int(attrs[0])
if f_num >= head and f_num <= seqL:
print >> outStream, line
outStream.close()
inStream.close()
return seqL
def createTxt(self, out_file):
f = open(out_file, 'w')
f.close()
def loadModel(self):
name = 'all_4'
self.Uphi = torch.load('Results/MOT16/IoU/%s/uphi_13.pth'%name).to(self.device)
self.Ephi = torch.load('Results/MOT16/IoU/%s/ephi_13.pth'%name).to(self.device)
self.u = torch.load('Results/MOT16/IoU/%s/u_13.pth'%name)
self.u = self.u.to(self.device)
def swapFC(self):
self.cur = self.cur ^ self.nxt
self.nxt = self.cur ^ self.nxt
self.cur = self.cur ^ self.nxt
def showCTau(self):
for i in xrange(2):
print ' Min:', self.ctau[i][0], 'Max:', self.ctau[i][1],
if self.ctau[i][3]:
print 'Mean:', self.ctau[i][2]/self.ctau[i][3]
else:
print 'Total:', self.ctau[i][2], 'Counter:', self.ctau[i][3]
print ' Overall mean:', (self.ctau[0][2]+self.ctau[1][2])/(self.ctau[0][3]+self.ctau[1][3])
def evaluation(self, head, tail, gtFile):
'''
Evaluation on dets
:param head: the head frame number
:param tail: the tail frame number
:param gtFile: the ground truth file name
:return: None
'''
gtIn = open(gtFile, 'r')
self.cur, self.nxt = 0, 1
line_con = [[], []]
id_step = 1
step = head + self.train_set.setBuffer(head)
while step < tail:
step += self.train_set.loadNext()
# print head+step, 'F',
u_ = self.Uphi(self.train_set.E, self.train_set.V, self.u)
# print 'Fo'
m = self.train_set.m
n = self.train_set.n
if n==0:
print 'There is no detection in the rest of sequence!'
break
if id_step == 1:
i = 0
while i < m:
attrs = gtIn.readline().strip().split(',')
if float(attrs[6]) >= tau_conf_score:
line_con[self.cur].append(attrs)
id_step += 1
i += 1
i = 0
while i < n:
attrs = gtIn.readline().strip().split(',')
if float(attrs[6]) >= tau_conf_score:
attrs.append(1)
line_con[self.nxt].append(attrs)
i += 1
# update the edges
# print 'T',
ret = self.train_set.getRet()
for edge in self.train_set.candidates:
e, vs_index, vr_index = edge
e = e.to(self.device).view(1,-1)
v1 = self.train_set.getMotion(1, vs_index).to(self.device)
v2 = self.train_set.getMotion(0, vr_index, vs_index, 1).to(self.device)
e_ = self.Ephi(e, v1, v2, u_)
self.train_set.edges[vs_index][vr_index] = e_.data.view(-1)
tmp = F.softmax(e_)
tmp = tmp.cpu().data.numpy()[0]
ret[vs_index][vr_index] = float(tmp[0])
# self.train_set.showE(outFile)
# for j in ret:
# print j
results = set(j for j in xrange(n))
for i in xrange(m):
a_attrs = line_con[self.cur][i]
for j in xrange(n):
index = 1
cost = ret[i][j]
# print a_attrs[1], line_con[self.nxt][j][1]
if a_attrs[1] == line_con[self.nxt][j][1]:
self.train_set.updateVelocity(i, j, 1)
results.remove(j)
index = 0
self.ctau[index][0] = min(self.ctau[index][0], cost)
self.ctau[index][1] = max(self.ctau[index][1], cost)
self.ctau[index][2] += cost
self.ctau[index][3] += 1
for j in results:
self.train_set.updateVelocity(-1, j)
line_con[self.cur] = []
# print head+step, results
self.train_set.swapFC()
self.swapFC()
gtIn.close()
# tra_tst = 'training sets' if head == 1 else 'validation sets'
# out = open(outFile, 'a')
# print >> out, tra_tst
# out.close()
self.showCTau()