def __init__(self, dim=512, n_obj_classes=151):
super(_RelPN, self).__init__()
self.anchor_scales = cfg.ANCHOR_SCALES
self.feat_stride = cfg.FEAT_STRIDE[0]
roi_feat_dim = n_obj_classes
if cfg.TRAIN.RELPN_WITH_BBOX_INFO:
roi_feat_dim += 4
self.RelPN_bilinear_sub = nn.Sequential(
nn.Linear(roi_feat_dim, 64),
nn.ReLU(inplace=True),
nn.Linear(64, 64),
)
self.RelPN_bilinear_obj = nn.Sequential(
nn.Linear(roi_feat_dim, 64),
nn.ReLU(inplace=True),
nn.Linear(64, 64),
)
# define proposal layer
self.RelPN_proposal = _ProposalLayer(self.feat_stride,
self.anchor_scales)
# define anchor target layer
self.RelPN_anchor_target = _AnchorTargetLayer(self.feat_stride,
self.anchor_scales)
self.relpn_loss_cls = 0
self.relpn_loss_box = 0
def __init__(self, din):
super(_RPN, self).__init__()
self.din = din # get depth of input feature map, e.g., 512
self.anchor_scales = cfg.ANCHOR_SCALES
self.anchor_ratios = cfg.ANCHOR_RATIOS
self.feat_stride = cfg.FEAT_STRIDE[0]
# define the convrelu layers processing input feature map
self.RPN_Conv = nn.Conv2d(self.din, 512, 3, 1, 1, bias=True)
# define bg/fg classifcation score layer
self.nc_score_out = len(self.anchor_scales) * len(self.anchor_ratios) * 2 # 2(bg/fg) * 9 (anchors)
self.RPN_cls_score = nn.Conv2d(512, self.nc_score_out, 1, 1, 0)
# define anchor box offset prediction layer
self.nc_bbox_out = len(self.anchor_scales) * len(self.anchor_ratios) * 4 # 4(coords) * 9 (anchors)
self.RPN_bbox_pred = nn.Conv2d(512, self.nc_bbox_out, 1, 1, 0)
# define proposal layer
self.RPN_proposal = _ProposalLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)
# define anchor target layer
self.RPN_anchor_target = _AnchorTargetLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)
self.rpn_loss_cls = 0
self.rpn_loss_box = 0
def __init__(self, anchors, all_anchors, inds_inside):
super(CPN, self).__init__()
self.image_shape = [[240, 320]
] # for one batch, TODO: maybe need to change here
self.anchors = anchors # (630, x, y, xw, yw) anchors coordinates
self.inds_inside = inds_inside
self.all_anchors = all_anchors
self.nc_score_out = 2 * 12 # 2(bg/fg) * 12 (anchors)
self.nc_bbox_out = 4 * 12 # 4(coords) * 12 (anchors)
c3d = C3D()
self.action_num = 22 #-1 # 21(classes) do not consider bg
self.action_anchor_num = self.action_num * 12
c3d.load_state_dict(torch.load(c3d_checkpoint))
self.RPN_Conv = nn.Conv2d(512, 512, 3, 1, 1, bias=True)
self.BN1 = nn.BatchNorm2d(512)
self.RPN_cls_bbox_action = nn.Conv2d(
512, self.nc_score_out + self.nc_bbox_out + self.action_anchor_num,
1, 1, 0)
self.BN2 = nn.BatchNorm2d(self.nc_score_out + self.nc_bbox_out +
self.action_anchor_num)
self.RPN_proposal = _ProposalLayer(self.anchors, self.all_anchors)
self.RPN_anchor_target = _AnchorTargetLayer(self.anchors,
self.inds_inside)