def _calc_ious(self, anchor, bbox, inside_index):
# ious between the anchors and the gt boxes
ious = bbox_iou(anchor, bbox)
argmax_ious = ious.argmax(axis=1)
max_ious = ious[np.arange(len(inside_index)), argmax_ious]
gt_argmax_ious = ious.argmax(axis=0)
gt_max_ious = ious[gt_argmax_ious, np.arange(ious.shape[1])]
gt_argmax_ious = np.where(ious == gt_max_ious)[0]
return argmax_ious, max_ious, gt_argmax_ious
def forward(self, gt_bboxes, gt_labels, nms_scores, sorted_labels, sorted_cls_bboxes):
sorted_score, prob_argsort = t.sort(nms_scores, descending=True)
sorted_cls_bboxes = sorted_cls_bboxes[prob_argsort]
sorted_labels = sorted_labels[prob_argsort]
sorted_labels = tonumpy(sorted_labels)
gt_labels = tonumpy(gt_labels)
nms_gt = t.zeros_like(sorted_score)
eps = 1e-8
iou = bbox_iou(tonumpy(gt_bboxes[0]), tonumpy(sorted_cls_bboxes))
for gt_idx in range(len(iou)):
accept_iou = np.reshape(np.argwhere(iou[gt_idx] > 0.5),-1)
accept_label = np.reshape(np.argwhere(sorted_labels[accept_iou] == gt_labels[0][gt_idx]),-1)
if not(len(accept_label)==0):
nms_gt[accept_iou[accept_label[0]]] = 1.
loss = nms_gt * (sorted_score+ eps).log() + (1 - nms_gt) * (1-sorted_score + eps).log()
loss = -loss.mean()
return loss
def calc_detection_voc_prec_rec(pred_bboxes,
pred_labels,
pred_scores,
gt_bboxes,
gt_labels,
gt_difficults=None,
iou_thresh=0.5):
"""Calculate precision and recall based on evaluation code of PASCAL VOC.
This function calculates precision and recall of
predicted bounding boxes obtained from a dataset which has :math:`N`
images.
The code is based on the evaluation code used in PASCAL VOC Challenge.
Args:
pred_bboxes (iterable of numpy.ndarray): An iterable of :math:`N`
sets of bounding boxes.
Its index corresponds to an index for the base dataset.
Each element of :obj:`pred_bboxes` is a set of coordinates
of bounding boxes. This is an array whose shape is :math:`(R, 4)`,
where :math:`R` corresponds
to the number of bounding boxes, which may vary among boxes.
The second axis corresponds to
:math:`y_{min}, x_{min}, y_{max}, x_{max}` of a bounding box.
pred_labels (iterable of numpy.ndarray): An iterable of labels.
Similar to :obj:`pred_bboxes`, its index corresponds to an
index for the base dataset. Its length is :math:`N`.
pred_scores (iterable of numpy.ndarray): An iterable of confidence
scores for predicted bounding boxes. Similar to :obj:`pred_bboxes`,
its index corresponds to an index for the base dataset.
Its length is :math:`N`.
gt_bboxes (iterable of numpy.ndarray): An iterable of ground truth
bounding boxes
whose length is :math:`N`. An element of :obj:`gt_bboxes` is a
bounding box whose shape is :math:`(R, 4)`. Note that the number of
bounding boxes in each image does not need to be same as the number
of corresponding predicted boxes.
gt_labels (iterable of numpy.ndarray): An iterable of ground truth
labels which are organized similarly to :obj:`gt_bboxes`.
gt_difficults (iterable of numpy.ndarray): An iterable of boolean
arrays which is organized similarly to :obj:`gt_bboxes`.
This tells whether the
corresponding ground truth bounding box is difficult or not.
By default, this is :obj:`None`. In that case, this function
considers all bounding boxes to be not difficult.
iou_thresh (float): A prediction is correct if its Intersection over
Union with the ground truth is above this value..
Returns:
tuple of two lists:
This function returns two lists: :obj:`prec` and :obj:`rec`.
* :obj:`prec`: A list of arrays. :obj:`prec[l]` is precision \
for class :math:`l`. If class :math:`l` does not exist in \
either :obj:`pred_labels` or :obj:`gt_labels`, :obj:`prec[l]` is \
set to :obj:`None`.
* :obj:`rec`: A list of arrays. :obj:`rec[l]` is recall \
for class :math:`l`. If class :math:`l` that is not marked as \
difficult does not exist in \
:obj:`gt_labels`, :obj:`rec[l]` is \
set to :obj:`None`.
"""
pred_bboxes = iter(pred_bboxes)
pred_labels = iter(pred_labels)
pred_scores = iter(pred_scores)
gt_bboxes = iter(gt_bboxes)
gt_labels = iter(gt_labels)
if gt_difficults is None:
gt_difficults = itertools.repeat(None)
else:
gt_difficults = iter(gt_difficults)
n_pos = defaultdict(int)
score = defaultdict(list)
match = defaultdict(list)
for pred_bbox, pred_label, pred_score, gt_bbox, gt_label, gt_difficult in \
six.moves.zip(
pred_bboxes, pred_labels, pred_scores,
gt_bboxes, gt_labels, gt_difficults):
if gt_difficult is None:
gt_difficult = np.zeros(gt_bbox.shape[0], dtype=bool)
for l in np.unique(np.concatenate((pred_label, gt_label)).astype(int)):
pred_mask_l = pred_label == l
pred_bbox_l = pred_bbox[pred_mask_l]
pred_score_l = pred_score[pred_mask_l]
# sort by score
order = pred_score_l.argsort()[::-1]
pred_bbox_l = pred_bbox_l[order]
pred_score_l = pred_score_l[order]
gt_mask_l = gt_label == l
gt_bbox_l = gt_bbox[gt_mask_l]
gt_difficult_l = gt_difficult[gt_mask_l]
n_pos[l] += np.logical_not(gt_difficult_l).sum()
score[l].extend(pred_score_l)
if len(pred_bbox_l) == 0:
continue
if len(gt_bbox_l) == 0:
match[l].extend((0, ) * pred_bbox_l.shape[0])
continue
# VOC evaluation follows integer typed bounding boxes.
pred_bbox_l = pred_bbox_l.copy()
pred_bbox_l[:, 2:] += 1
gt_bbox_l = gt_bbox_l.copy()
gt_bbox_l[:, 2:] += 1
iou = bbox_iou(pred_bbox_l, gt_bbox_l)
gt_index = iou.argmax(axis=1)
# set -1 if there is no matching ground truth
gt_index[iou.max(axis=1) < iou_thresh] = -1
del iou
selec = np.zeros(gt_bbox_l.shape[0], dtype=bool)
for gt_idx in gt_index:
if gt_idx >= 0:
if gt_difficult_l[gt_idx]:
match[l].append(-1)
else:
if not selec[gt_idx]:
match[l].append(1)
else:
match[l].append(0)
selec[gt_idx] = True
else:
match[l].append(0)
for iter_ in (pred_bboxes, pred_labels, pred_scores, gt_bboxes, gt_labels,
gt_difficults):
if next(iter_, None) is not None:
raise ValueError('Length of input iterables need to be same.')
n_fg_class = max(n_pos.keys()) + 1
prec = [None] * n_fg_class
rec = [None] * n_fg_class
for l in n_pos.keys():
score_l = np.array(score[l])
match_l = np.array(match[l], dtype=np.int8)
order = score_l.argsort()[::-1]
match_l = match_l[order]
tp = np.cumsum(match_l == 1)
fp = np.cumsum(match_l == 0)
# If an element of fp + tp is 0,
# the corresponding element of prec[l] is nan.
prec[l] = tp / (fp + tp)
# If n_pos[l] is 0, rec[l] is None.
if n_pos[l] > 0:
rec[l] = tp / n_pos[l]
return prec, rec
def __call__(self, roi, bbox, label,
loc_normalize_mean=(0., 0., 0., 0.),
loc_normalize_std=(0.1, 0.1, 0.2, 0.2)):
"""Assigns ground truth to sampled proposals.
This function samples total of :obj:`self.n_sample` RoIs
from the combination of :obj:`roi` and :obj:`bbox`.
The RoIs are assigned with the ground truth class labels as well as
bounding box offsets and scales to match the ground truth bounding
boxes. As many as :obj:`pos_ratio * self.n_sample` RoIs are
sampled as foregrounds.
Offsets and scales of bounding boxes are calculated using
:func:`model.utils.bbox_tools.bbox2loc`.
Also, types of input arrays and output arrays are same.
Here are notations.
* :math:`S` is the total number of sampled RoIs, which equals \
:obj:`self.n_sample`.
* :math:`L` is number of object classes possibly including the \
background.
Args:
roi (array): Region of Interests (RoIs) from which we sample.
Its shape is :math:`(R, 4)`
bbox (array): The coordinates of ground truth bounding boxes.
Its shape is :math:`(R', 4)`.
label (array): Ground truth bounding box labels. Its shape
is :math:`(R',)`. Its range is :math:`[0, L - 1]`, where
:math:`L` is the number of foreground classes.
loc_normalize_mean (tuple of four floats): Mean values to normalize
coordinates of bouding boxes.
loc_normalize_std (tupler of four floats): Standard deviation of
the coordinates of bounding boxes.
Returns:
(array, array, array):
* **sample_roi**: Regions of interests that are sampled. \
Its shape is :math:`(S, 4)`.
* **gt_roi_loc**: Offsets and scales to match \
the sampled RoIs to the ground truth bounding boxes. \
Its shape is :math:`(S, 4)`.
* **gt_roi_label**: Labels assigned to sampled RoIs. Its shape is \
:math:`(S,)`. Its range is :math:`[0, L]`. The label with \
value 0 is the background.
"""
n_bbox, _ = bbox.shape
roi = np.concatenate((roi, bbox), axis=0)
pos_roi_per_image = np.round(self.n_sample * self.pos_ratio)
iou = bbox_iou(roi, bbox)
gt_assignment = iou.argmax(axis=1)
max_iou = iou.max(axis=1)
# Offset range of classes from [0, n_fg_class - 1] to [1, n_fg_class].
# The label with value 0 is the background.
gt_roi_label = label[gt_assignment] + 1
# Select foreground RoIs as those with >= pos_iou_thresh IoU.
pos_index = np.where(max_iou >= self.pos_iou_thresh)[0]
pos_roi_per_this_image = int(min(pos_roi_per_image, pos_index.size))
if pos_index.size > 0:
pos_index = np.random.choice(
pos_index, size=pos_roi_per_this_image, replace=False)
# Select background RoIs as those within
# [neg_iou_thresh_lo, neg_iou_thresh_hi).
neg_index = np.where((max_iou < self.neg_iou_thresh_hi) &
(max_iou >= self.neg_iou_thresh_lo))[0]
neg_roi_per_this_image = self.n_sample - pos_roi_per_this_image
neg_roi_per_this_image = int(min(neg_roi_per_this_image,
neg_index.size))
if neg_index.size > 0:
neg_index = np.random.choice(
neg_index, size=neg_roi_per_this_image, replace=False)
# The indices that we're selecting (both positive and negative).
keep_index = np.append(pos_index, neg_index)
gt_roi_label = gt_roi_label[keep_index]
gt_roi_label[pos_roi_per_this_image:] = 0 # negative labels --> 0
sample_roi = roi[keep_index]
# Compute offsets and scales to match sampled RoIs to the GTs.
gt_roi_loc = bbox2loc(sample_roi, bbox[gt_assignment[keep_index]])
gt_roi_loc = ((gt_roi_loc - np.array(loc_normalize_mean, np.float32)
) / np.array(loc_normalize_std, np.float32))
return sample_roi, gt_roi_loc, gt_roi_label