def proposal_layer(rpn_cls_prob_reshape, rpn_bbox_pred, im_info, cfg_key,
_feat_stride, anchor_scales):
rpn_cls_prob_reshape = rpn_cls_prob_reshape.data.cpu().numpy()
rpn_bbox_pred = rpn_bbox_pred.data.cpu().numpy()
x = proposal_layer_py(rpn_cls_prob_reshape, rpn_bbox_pred, im_info,
cfg_key, _feat_stride, anchor_scales)
x = faster_rcnn.network.np_to_variable(x, is_cuda=True)
return x.view(-1, 5)
def proposal_layer(rpn_cls_prob_reshape, rpn_bbox_pred, im_info, cfg_key,
_feat_stride, anchor_scales):
"""
:param rpn_cls_prob_reshape: (1, 2*9_anchors, 37, 50) tensor
:param rpn_bbox_pred: (1, 4*9_anchors, 37, 50) tensor
:param im_info: (1, 3) numpy
:param cfg_key:
:param _feat_stride:
:param anchor_scales:
:return:
"""
rpn_cls_prob_reshape = rpn_cls_prob_reshape.data.cpu().numpy()
rpn_bbox_pred = rpn_bbox_pred.data.cpu().numpy()
x = proposal_layer_py(rpn_cls_prob_reshape, rpn_bbox_pred, im_info,
cfg_key, _feat_stride, anchor_scales)
x = network.np_to_variable(x, is_cuda=True)
return x.view(-1, 5)
def __init__(self, debug=False):
super(RPN, self).__init__()
self.anchor_scales = cfg.ANCHOR_SCALES
self.anchor_ratios = cfg.ANCHOR_RATIOS
self.feat_stride = cfg.FEAT_STRIDE[0]
self._vgg16 = VGG16()
self.conv1 = Conv2d(512, 512, 3, same_padding=True)
self.score_conv = Conv2d(512,
len(self.anchor_scales) *
len(self.anchor_ratios) * 2,
1,
relu=False,
same_padding=False)
self.bbox_conv = Conv2d(512,
len(self.anchor_scales) *
len(self.anchor_ratios) * 4,
1,
relu=False,
same_padding=False)
# define proposal layer
self.proposal_layer = proposal_layer_py(self.feat_stride,
self.anchor_scales,
self.anchor_ratios)
# define anchor target layer
self.anchor_target_layer = anchor_target_layer_py(
self.feat_stride, self.anchor_scales, self.anchor_ratios)
# loss
self.cross_entropy = 0
self.loss_box = 0
# for log
self.debug = debug