def forward(self, predictions):
"""
Args:
loc_data: (tensor) Loc preds from loc layers
Shape: [batch,num_priors*4]
conf_data: (tensor) Shape: Conf preds from conf layers
Shape: [batch*num_priors,num_classes]
prior_data: (tensor) Prior boxes and variances from priorbox layers
Shape: [1,num_priors,4]
"""
loc, conf = predictions
loc_data = loc.data
conf_data = conf.data
prior_data = self.priors.data
num = loc_data.size(0) # batch size
num_priors = prior_data.size(0)
output = torch.zeros(num, self.num_classes, self.top_k, 5)
conf_preds = conf_data.view(num, num_priors,
self.num_classes).transpose(2, 1)
# Decode predictions into bboxes.
for i in range(num):
decoded_boxes = decode(loc_data[i], prior_data, self.variance)
# For each class, perform nms
conf_scores = conf_preds[i].clone()
for cl in range(1, self.num_classes):
c_mask = conf_scores[cl].gt(
self.conf_thresh).nonzero().view(-1)
if c_mask.dim() == 0:
continue
scores = conf_scores[cl][c_mask]
if scores.size(0) == 0 or scores.dim() == 0:
continue
# l_mask = c_mask.unsqueeze(1).expand_as(decoded_boxes)
# boxes = decoded_boxes[l_mask].view(-1, 4)
boxes = decoded_boxes[c_mask, :]
# idx of highest scoring and non-overlapping boxes per class
ids, count = nms(boxes, scores, self.nms_thresh, self.top_k)
output[i, cl, :count] = \
torch.cat((scores[ids[:count]].unsqueeze(1),
boxes[ids[:count]]), 1)
return output
def forward(self, predictions):
"""
Args:
loc_data: (tensor) Loc preds from loc layers
Shape: [batch,num_priors*4]
conf_data: (tensor) Shape: Conf preds from conf layers
Shape: [batch*num_priors,num_classes]
prior_data: (tensor) Prior boxes and variances from priorbox layers
Shape: [1,num_priors,4]
"""
loc, conf = predictions
loc_data = loc.data
conf_data = conf.data
prior_data = self.priors.data
num = loc_data.size(0) # batch size
num_priors = prior_data.size(0)
#self.output.zero_()
if num == 1:
# size batch x num_classes x num_priors
conf_preds = conf_data.t().contiguous().unsqueeze(0)
else:
conf_preds = conf_data.view(num, num_priors,
self.num_classes).transpose(2, 1)
#self.output.expand_(num, self.num_classes, self.top_k, 5)
output = torch.zeros(num, self.num_classes, self.top_k, 5)
_t = {'decode': Timer(), 'misc': Timer(), 'box_mask':Timer(), 'score_mask':Timer(),'nms':Timer(), 'cpu':Timer(),'sort':Timer()}
gpunms_time = 0
scores_time=0
box_time=0
cpu_tims=0
sort_time=0
decode_time=0
_t['misc'].tic()
# Decode predictions into bboxes.
for i in range(num):
_t['decode'].tic()
decoded_boxes = decode(loc_data[i], prior_data, self.variance)
decode_time += _t['decode'].toc()
# For each class, perform nms
conf_scores = conf_preds[i].clone()
num_det = 0
for cl in range(1, self.num_classes):
_t['cpu'].tic()
c_mask = conf_scores[cl].gt(self.conf_thresh).nonzero().view(-1)
cpu_tims+=_t['cpu'].toc()
if c_mask.dim() == 0:
continue
_t['score_mask'].tic()
scores = conf_scores[cl][c_mask]
scores_time+=_t['score_mask'].toc()
if scores.dim() == 0:
continue
_t['box_mask'].tic()
# l_mask = c_mask.unsqueeze(1).expand_as(decoded_boxes)
# boxes = decoded_boxes[l_mask].view(-1, 4)
boxes = decoded_boxes[c_mask, :]
box_time+=_t['box_mask'].toc()
# idx of highest scoring and non-overlapping boxes per class
_t['nms'].tic()
# cls_dets = torch.cat((boxes, scores), 1)
# _, order = torch.sort(scores, 0, True)
# cls_dets = cls_dets[order]
# keep = nms(cls_dets, self.nms_thresh)
# cls_dets = cls_dets[keep.view(-1).long()]
ids, count = nms(boxes, scores, self.nms_thresh, self.top_k)
gpunms_time += _t['nms'].toc()
output[i, cl, :count] = \
torch.cat((scores[ids[:count]].unsqueeze(1),
boxes[ids[:count]]), 1)
nms_time= _t['misc'].toc()
# print(nms_time, cpu_tims, scores_time,box_time,gpunms_time)
# flt = self.output.view(-1, 5)
# _, idx = flt[:, 0].sort(0)
# _, rank = idx.sort(0)
# flt[(rank >= self.top_k).unsqueeze(1).expand_as(flt)].fill_(0)
return output
def forward(self, predictions, return_time=False):
"""
Args:
loc_data: (tensor) Loc preds from loc layers
Shape: [batch,num_priors*4]
conf_data: (tensor) Shape: Conf preds from conf layers
Shape: [batch*num_priors,num_classes]
prior_data: (tensor) Prior boxes and variances from priorbox layers
Shape: [1,num_priors,4]
"""
loc, conf = predictions
loc_data = loc.data
conf_data = conf.data
prior_data = self.priors.data
num = loc_data.size(0) # batch size
num_priors = prior_data.size(0)
#self.output.zero_()
if num == 1:
# size batch x num_classes x num_priors
conf_preds = conf_data.t().contiguous().unsqueeze(0)
else:
conf_preds = conf_data.view(num, num_priors,
self.num_classes).transpose(2, 1)
#self.output.expand_(num, self.num_classes, self.top_k, 5)
output = torch.zeros(num, self.num_classes, self.top_k, 5)
_t = {'decode': Timer(), 'misc': Timer(), 'box_mask':Timer(), 'score_mask':Timer(),'nms':Timer(), 'cpu':Timer(),'sort':Timer()}
from collections import defaultdict
time_taken = defaultdict(float)
_t['misc'].tic()
# Decode predictions into bboxes.
for i in range(num):
_t['decode'].tic()
decoded_boxes = decode(loc_data[i], prior_data, self.variance)
time_taken['decode'] += _t['decode'].toc()
# For each class, perform nms
conf_scores = conf_preds[i].clone()
num_det = 0
for cl in range(1, self.num_classes):
_t['cpu'].tic()
c_mask = conf_scores[cl].gt(self.conf_thresh).nonzero().view(-1)
time_taken['cpu'] +=_t['cpu'].toc()
if c_mask.dim() == 0:
continue
_t['score_mask'].tic()
scores = conf_scores[cl][c_mask]
time_taken['score_mask'] +=_t['score_mask'].toc()
if scores.dim() == 0:
continue
_t['box_mask'].tic()
# l_mask = c_mask.unsqueeze(1).expand_as(decoded_boxes)
# boxes = decoded_boxes[l_mask].view(-1, 4)
boxes = decoded_boxes[c_mask, :]
time_taken['box_mask'] +=_t['box_mask'].toc()
# idx of highest scoring and non-overlapping boxes per class
_t['nms'].tic()
# cls_dets = torch.cat((boxes, scores), 1)
# _, order = torch.sort(scores, 0, True)
# cls_dets = cls_dets[order]
# keep = nms(cls_dets, self.nms_thresh)
# cls_dets = cls_dets[keep.view(-1).long()]
try:
ids, count = nms(boxes, scores, self.nms_thresh, self.top_k)
except Error as e:
logger.error(e)
logger.debug("BOXES: \n{} \nSCORES: {}\nTHRESH: {}\nTOP_K: {}".format(
boxes, scores, self.nms_thresh, self.top_k))
time_taken['gpu_nms'] += _t['nms'].toc()
output[i, cl, :count] = \
torch.cat((scores[ids[:count]].unsqueeze(1), boxes[ids[:count]]), 1) \
time_taken['nms_total'] = _t['misc'].toc()
#keys = time_taken.keys()
#time_str_format = "{}: {:.4f}s | "
#str_stats = 'NMS TIME TAKEN - ' + ''.join([
# time_str_format.format(k, time_taken[k])
# for k in ['cpu', 'gpu_nms', 'nms_total']
#])
#logger.debug(str_stats)
# print(nms_time, cpu_tims, scores_time,box_time,gpunms_time)
# flt = self.output.view(-1, 5)
# _, idx = flt[:, 0].sort(0)
# _, rank = idx.sort(0)
# flt[(rank >= self.top_k).unsqueeze(1).expand_as(flt)].fill_(0)
if return_time:
return output, time_taken
else:
return output
def forward(self, predictions):
"""
Args:
loc_data: (tensor) Loc preds from loc layers
Shape: [batch,num_priors*4]
conf_data: (tensor) Shape: Conf preds from conf layers
Shape: [batch*num_priors,num_classes]
prior_data: (tensor) Prior boxes and variances from priorbox layers
Shape: [1,num_priors,4]
"""
loc, conf = predictions
loc_data = loc.data
conf_data = conf.data
prior_data = self.priors.data
num = loc_data.size(0) # batch size
num_priors = prior_data.size(0)
# self.output.zero_()
if num == 1:
# size batch x num_classes x num_priors
conf_preds = conf_data.t().contiguous().unsqueeze(0)
else:
conf_preds = conf_data.view(num, num_priors,
self.num_classes).transpose(2, 1)
# self.output.expand_(num, self.num_classes, self.top_k, 5)
output = torch.zeros(num, self.num_classes, self.top_k, 5)
# Decode predictions into bboxes.
for i in range(num):
decoded_boxes = decode(loc_data[i], prior_data, self.variance)
# print('decoded_boxes: {}'.format(decoded_boxes))
# For each class, perform nms
conf_scores = conf_preds[i].clone()
# print('conf_scores: {}'.format(conf_scores))
# print(conf_scores.size())
for cl in range(1, self.num_classes):
# print(conf_scores[cl])
# print(conf_scores[cl].size())
c_mask = conf_scores[cl].gt(
self.conf_thresh).nonzero().view(-1)
# print('cmask: ', c_mask)
if c_mask.dim() == 0:
continue
scores = conf_scores[cl][c_mask]
if scores.dim() == 0:
continue
# l_mask = c_mask.unsqueeze(1).expand_as(decoded_boxes)
# boxes = decoded_boxes[l_mask].view(-1, 4)
boxes = decoded_boxes[c_mask, :]
# print(scores, boxes)
# idx of highest scoring and non-overlapping boxes per class
# cls_dets = torch.cat((boxes, scores), 1)
# _, order = torch.sort(scores, 0, True)
# cls_dets = cls_dets[order]
# keep = nms(cls_dets, self.nms_thresh)
# cls_dets = cls_dets[keep.view(-1).long()]
# print('before nms:')
# print('boxes: {}'.format(boxes))
# print('scores: {}'.format(scores))
# why it is empty?
ids, count = nms(boxes, scores, self.nms_thresh, self.top_k)
output[i, cl, :count] = \
torch.cat((scores[ids[:count]].unsqueeze(1),
boxes[ids[:count]]), 1)
# print(nms_time, cpu_tims, scores_time,box_time,gpunms_time)
# flt = self.output.view(-1, 5)
# _, idx = flt[:, 0].sort(0)
# _, rank = idx.sort(0)
# flt[(rank >= self.top_k).unsqueeze(1).expand_as(flt)].fill_(0)
return output