import torch
import torch.nn as nn
from mmcv.cnn import ConvModule, Scale, bias_init_with_prob, normal_init
from mmdet.core import distance2bbox, force_fp32, multi_apply, multiclass_nms
from ..builder import HEADS, build_loss
INF = 1e8
[docs]@HEADS.register_module()
class FCOSHead(nn.Module):
"""Anchor-free head used in `FCOS <https://arxiv.org/abs/1904.01355>`_.
The FCOS head does not use anchor boxes. Instead bounding boxes are
predicted at each pixel and a centerness measure is used to supress
low-quality predictions.
Example:
>>> self = FCOSHead(11, 7)
>>> feats = [torch.rand(1, 7, s, s) for s in [4, 8, 16, 32, 64]]
>>> cls_score, bbox_pred, centerness = self.forward(feats)
>>> assert len(cls_score) == len(self.scales)
"""
def __init__(self,
num_classes,
in_channels,
feat_channels=256,
stacked_convs=4,
strides=(4, 8, 16, 32, 64),
regress_ranges=((-1, 64), (64, 128), (128, 256), (256, 512),
(512, INF)),
center_sampling=False,
center_sample_radius=1.5,
background_label=None,
loss_cls=dict(
type='FocalLoss',
use_sigmoid=True,
gamma=2.0,
alpha=0.25,
loss_weight=1.0),
loss_bbox=dict(type='IoULoss', loss_weight=1.0),
loss_centerness=dict(
type='CrossEntropyLoss',
use_sigmoid=True,
loss_weight=1.0),
conv_cfg=None,
norm_cfg=dict(type='GN', num_groups=32, requires_grad=True),
train_cfg=None,
test_cfg=None):
super(FCOSHead, self).__init__()
self.num_classes = num_classes
self.cls_out_channels = num_classes
self.in_channels = in_channels
self.feat_channels = feat_channels
self.stacked_convs = stacked_convs
self.strides = strides
self.regress_ranges = regress_ranges
self.loss_cls = build_loss(loss_cls)
self.loss_bbox = build_loss(loss_bbox)
self.loss_centerness = build_loss(loss_centerness)
self.train_cfg = train_cfg
self.test_cfg = test_cfg
self.conv_cfg = conv_cfg
self.norm_cfg = norm_cfg
self.fp16_enabled = False
self.center_sampling = center_sampling
self.center_sample_radius = center_sample_radius
self.background_label = (
num_classes if background_label is None else background_label)
# background_label should be either 0 or num_classes
assert (self.background_label == 0
or self.background_label == num_classes)
self._init_layers()
def _init_layers(self):
self.cls_convs = nn.ModuleList()
self.reg_convs = nn.ModuleList()
for i in range(self.stacked_convs):
chn = self.in_channels if i == 0 else self.feat_channels
self.cls_convs.append(
ConvModule(
chn,
self.feat_channels,
3,
stride=1,
padding=1,
conv_cfg=self.conv_cfg,
norm_cfg=self.norm_cfg,
bias=self.norm_cfg is None))
self.reg_convs.append(
ConvModule(
chn,
self.feat_channels,
3,
stride=1,
padding=1,
conv_cfg=self.conv_cfg,
norm_cfg=self.norm_cfg,
bias=self.norm_cfg is None))
self.fcos_cls = nn.Conv2d(
self.feat_channels, self.cls_out_channels, 3, padding=1)
self.fcos_reg = nn.Conv2d(self.feat_channels, 4, 3, padding=1)
self.fcos_centerness = nn.Conv2d(self.feat_channels, 1, 3, padding=1)
self.scales = nn.ModuleList([Scale(1.0) for _ in self.strides])
def init_weights(self):
for m in self.cls_convs:
normal_init(m.conv, std=0.01)
for m in self.reg_convs:
normal_init(m.conv, std=0.01)
bias_cls = bias_init_with_prob(0.01)
normal_init(self.fcos_cls, std=0.01, bias=bias_cls)
normal_init(self.fcos_reg, std=0.01)
normal_init(self.fcos_centerness, std=0.01)
[docs] def forward(self, feats):
return multi_apply(self.forward_single, feats, self.scales)
def forward_single(self, x, scale):
cls_feat = x
reg_feat = x
for cls_layer in self.cls_convs:
cls_feat = cls_layer(cls_feat)
cls_score = self.fcos_cls(cls_feat)
centerness = self.fcos_centerness(cls_feat)
for reg_layer in self.reg_convs:
reg_feat = reg_layer(reg_feat)
# scale the bbox_pred of different level
# float to avoid overflow when enabling FP16
bbox_pred = scale(self.fcos_reg(reg_feat)).float().exp()
return cls_score, bbox_pred, centerness
@force_fp32(apply_to=('cls_scores', 'bbox_preds', 'centernesses'))
def loss(self,
cls_scores,
bbox_preds,
centernesses,
gt_bboxes,
gt_labels,
img_metas,
gt_bboxes_ignore=None):
assert len(cls_scores) == len(bbox_preds) == len(centernesses)
featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
all_level_points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
bbox_preds[0].device)
labels, bbox_targets = self.get_targets(all_level_points, gt_bboxes,
gt_labels)
num_imgs = cls_scores[0].size(0)
# flatten cls_scores, bbox_preds and centerness
flatten_cls_scores = [
cls_score.permute(0, 2, 3, 1).reshape(-1, self.cls_out_channels)
for cls_score in cls_scores
]
flatten_bbox_preds = [
bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4)
for bbox_pred in bbox_preds
]
flatten_centerness = [
centerness.permute(0, 2, 3, 1).reshape(-1)
for centerness in centernesses
]
flatten_cls_scores = torch.cat(flatten_cls_scores)
flatten_bbox_preds = torch.cat(flatten_bbox_preds)
flatten_centerness = torch.cat(flatten_centerness)
flatten_labels = torch.cat(labels)
flatten_bbox_targets = torch.cat(bbox_targets)
# repeat points to align with bbox_preds
flatten_points = torch.cat(
[points.repeat(num_imgs, 1) for points in all_level_points])
# FG cat_id: [0, num_classes -1], BG cat_id: num_classes
bg_class_ind = self.num_classes
pos_inds = ((flatten_labels >= 0)
& (flatten_labels < bg_class_ind)).nonzero().reshape(-1)
num_pos = len(pos_inds)
loss_cls = self.loss_cls(
flatten_cls_scores, flatten_labels,
avg_factor=num_pos + num_imgs) # avoid num_pos is 0
pos_bbox_preds = flatten_bbox_preds[pos_inds]
pos_centerness = flatten_centerness[pos_inds]
if num_pos > 0:
pos_bbox_targets = flatten_bbox_targets[pos_inds]
pos_centerness_targets = self.centerness_target(pos_bbox_targets)
pos_points = flatten_points[pos_inds]
pos_decoded_bbox_preds = distance2bbox(pos_points, pos_bbox_preds)
pos_decoded_target_preds = distance2bbox(pos_points,
pos_bbox_targets)
# centerness weighted iou loss
loss_bbox = self.loss_bbox(
pos_decoded_bbox_preds,
pos_decoded_target_preds,
weight=pos_centerness_targets,
avg_factor=pos_centerness_targets.sum())
loss_centerness = self.loss_centerness(pos_centerness,
pos_centerness_targets)
else:
loss_bbox = pos_bbox_preds.sum()
loss_centerness = pos_centerness.sum()
return dict(
loss_cls=loss_cls,
loss_bbox=loss_bbox,
loss_centerness=loss_centerness)
@force_fp32(apply_to=('cls_scores', 'bbox_preds', 'centernesses'))
def get_bboxes(self,
cls_scores,
bbox_preds,
centernesses,
img_metas,
cfg=None,
rescale=None):
assert len(cls_scores) == len(bbox_preds)
num_levels = len(cls_scores)
featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
mlvl_points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
bbox_preds[0].device)
result_list = []
for img_id in range(len(img_metas)):
cls_score_list = [
cls_scores[i][img_id].detach() for i in range(num_levels)
]
bbox_pred_list = [
bbox_preds[i][img_id].detach() for i in range(num_levels)
]
centerness_pred_list = [
centernesses[i][img_id].detach() for i in range(num_levels)
]
img_shape = img_metas[img_id]['img_shape']
scale_factor = img_metas[img_id]['scale_factor']
det_bboxes = self._get_bboxes_single(cls_score_list,
bbox_pred_list,
centerness_pred_list,
mlvl_points, img_shape,
scale_factor, cfg, rescale)
result_list.append(det_bboxes)
return result_list
def _get_bboxes_single(self,
cls_scores,
bbox_preds,
centernesses,
mlvl_points,
img_shape,
scale_factor,
cfg,
rescale=False):
cfg = self.test_cfg if cfg is None else cfg
assert len(cls_scores) == len(bbox_preds) == len(mlvl_points)
mlvl_bboxes = []
mlvl_scores = []
mlvl_centerness = []
for cls_score, bbox_pred, centerness, points in zip(
cls_scores, bbox_preds, centernesses, mlvl_points):
assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
scores = cls_score.permute(1, 2, 0).reshape(
-1, self.cls_out_channels).sigmoid()
centerness = centerness.permute(1, 2, 0).reshape(-1).sigmoid()
bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4)
nms_pre = cfg.get('nms_pre', -1)
if nms_pre > 0 and scores.shape[0] > nms_pre:
max_scores, _ = (scores * centerness[:, None]).max(dim=1)
_, topk_inds = max_scores.topk(nms_pre)
points = points[topk_inds, :]
bbox_pred = bbox_pred[topk_inds, :]
scores = scores[topk_inds, :]
centerness = centerness[topk_inds]
bboxes = distance2bbox(points, bbox_pred, max_shape=img_shape)
mlvl_bboxes.append(bboxes)
mlvl_scores.append(scores)
mlvl_centerness.append(centerness)
mlvl_bboxes = torch.cat(mlvl_bboxes)
if rescale:
mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
mlvl_scores = torch.cat(mlvl_scores)
padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
# remind that we set FG labels to [0, num_class-1] since mmdet v2.0
# BG cat_id: num_class
mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
mlvl_centerness = torch.cat(mlvl_centerness)
det_bboxes, det_labels = multiclass_nms(
mlvl_bboxes,
mlvl_scores,
cfg.score_thr,
cfg.nms,
cfg.max_per_img,
score_factors=mlvl_centerness)
return det_bboxes, det_labels
[docs] def get_points(self, featmap_sizes, dtype, device):
"""Get points according to feature map sizes.
Args:
featmap_sizes (list[tuple]): Multi-level feature map sizes.
dtype (torch.dtype): Type of points.
device (torch.device): Device of points.
Returns:
tuple: points of each image.
"""
mlvl_points = []
for i in range(len(featmap_sizes)):
mlvl_points.append(
self._get_points_single(featmap_sizes[i], self.strides[i],
dtype, device))
return mlvl_points
def _get_points_single(self, featmap_size, stride, dtype, device):
h, w = featmap_size
x_range = torch.arange(
0, w * stride, stride, dtype=dtype, device=device)
y_range = torch.arange(
0, h * stride, stride, dtype=dtype, device=device)
y, x = torch.meshgrid(y_range, x_range)
points = torch.stack(
(x.reshape(-1), y.reshape(-1)), dim=-1) + stride // 2
return points
def get_targets(self, points, gt_bboxes_list, gt_labels_list):
assert len(points) == len(self.regress_ranges)
num_levels = len(points)
# expand regress ranges to align with points
expanded_regress_ranges = [
points[i].new_tensor(self.regress_ranges[i])[None].expand_as(
points[i]) for i in range(num_levels)
]
# concat all levels points and regress ranges
concat_regress_ranges = torch.cat(expanded_regress_ranges, dim=0)
concat_points = torch.cat(points, dim=0)
# the number of points per img, per lvl
num_points = [center.size(0) for center in points]
# get labels and bbox_targets of each image
labels_list, bbox_targets_list = multi_apply(
self._get_target_single,
gt_bboxes_list,
gt_labels_list,
points=concat_points,
regress_ranges=concat_regress_ranges,
num_points_per_lvl=num_points)
# split to per img, per level
labels_list = [labels.split(num_points, 0) for labels in labels_list]
bbox_targets_list = [
bbox_targets.split(num_points, 0)
for bbox_targets in bbox_targets_list
]
# concat per level image
concat_lvl_labels = []
concat_lvl_bbox_targets = []
for i in range(num_levels):
concat_lvl_labels.append(
torch.cat([labels[i] for labels in labels_list]))
concat_lvl_bbox_targets.append(
torch.cat(
[bbox_targets[i] for bbox_targets in bbox_targets_list]))
return concat_lvl_labels, concat_lvl_bbox_targets
def _get_target_single(self, gt_bboxes, gt_labels, points, regress_ranges,
num_points_per_lvl):
num_points = points.size(0)
num_gts = gt_labels.size(0)
if num_gts == 0:
return gt_labels.new_full((num_points, ), self.background_label), \
gt_bboxes.new_zeros((num_points, 4))
areas = (gt_bboxes[:, 2] - gt_bboxes[:, 0]) * (
gt_bboxes[:, 3] - gt_bboxes[:, 1])
# TODO: figure out why these two are different
# areas = areas[None].expand(num_points, num_gts)
areas = areas[None].repeat(num_points, 1)
regress_ranges = regress_ranges[:, None, :].expand(
num_points, num_gts, 2)
gt_bboxes = gt_bboxes[None].expand(num_points, num_gts, 4)
xs, ys = points[:, 0], points[:, 1]
xs = xs[:, None].expand(num_points, num_gts)
ys = ys[:, None].expand(num_points, num_gts)
left = xs - gt_bboxes[..., 0]
right = gt_bboxes[..., 2] - xs
top = ys - gt_bboxes[..., 1]
bottom = gt_bboxes[..., 3] - ys
bbox_targets = torch.stack((left, top, right, bottom), -1)
if self.center_sampling:
# condition1: inside a `center bbox`
radius = self.center_sample_radius
center_xs = (gt_bboxes[..., 0] + gt_bboxes[..., 2]) / 2
center_ys = (gt_bboxes[..., 1] + gt_bboxes[..., 3]) / 2
center_gts = torch.zeros_like(gt_bboxes)
stride = center_xs.new_zeros(center_xs.shape)
# project the points on current lvl back to the `original` sizes
lvl_begin = 0
for lvl_idx, num_points_lvl in enumerate(num_points_per_lvl):
lvl_end = lvl_begin + num_points_lvl
stride[lvl_begin:lvl_end] = self.strides[lvl_idx] * radius
lvl_begin = lvl_end
x_mins = center_xs - stride
y_mins = center_ys - stride
x_maxs = center_xs + stride
y_maxs = center_ys + stride
center_gts[..., 0] = torch.where(x_mins > gt_bboxes[..., 0],
x_mins, gt_bboxes[..., 0])
center_gts[..., 1] = torch.where(y_mins > gt_bboxes[..., 1],
y_mins, gt_bboxes[..., 1])
center_gts[..., 2] = torch.where(x_maxs > gt_bboxes[..., 2],
gt_bboxes[..., 2], x_maxs)
center_gts[..., 3] = torch.where(y_maxs > gt_bboxes[..., 3],
gt_bboxes[..., 3], y_maxs)
cb_dist_left = xs - center_gts[..., 0]
cb_dist_right = center_gts[..., 2] - xs
cb_dist_top = ys - center_gts[..., 1]
cb_dist_bottom = center_gts[..., 3] - ys
center_bbox = torch.stack(
(cb_dist_left, cb_dist_top, cb_dist_right, cb_dist_bottom), -1)
inside_gt_bbox_mask = center_bbox.min(-1)[0] > 0
else:
# condition1: inside a gt bbox
inside_gt_bbox_mask = bbox_targets.min(-1)[0] > 0
# condition2: limit the regression range for each location
max_regress_distance = bbox_targets.max(-1)[0]
inside_regress_range = (
max_regress_distance >= regress_ranges[..., 0]) & (
max_regress_distance <= regress_ranges[..., 1])
# if there are still more than one objects for a location,
# we choose the one with minimal area
areas[inside_gt_bbox_mask == 0] = INF
areas[inside_regress_range == 0] = INF
min_area, min_area_inds = areas.min(dim=1)
labels = gt_labels[min_area_inds]
labels[min_area == INF] = self.background_label # set as BG
bbox_targets = bbox_targets[range(num_points), min_area_inds]
return labels, bbox_targets
def centerness_target(self, pos_bbox_targets):
# only calculate pos centerness targets, otherwise there may be nan
left_right = pos_bbox_targets[:, [0, 2]]
top_bottom = pos_bbox_targets[:, [1, 3]]
centerness_targets = (
left_right.min(dim=-1)[0] / left_right.max(dim=-1)[0]) * (
top_bottom.min(dim=-1)[0] / top_bottom.max(dim=-1)[0])
return torch.sqrt(centerness_targets)