Python fast_rcnn.config.cfg.EPS Examples

def _compute_targets(ex_rois, gt_rois):
    """Compute bounding-box regression targets for an image. The targets are scale invariance"""

    ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + cfg.EPS
    ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + cfg.EPS
    ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths
    ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights

    gt_widths = gt_rois[:, 2] - gt_rois[:, 0] + cfg.EPS
    gt_heights = gt_rois[:, 3] - gt_rois[:, 1] + cfg.EPS
    gt_ctr_x = gt_rois[:, 0] + 0.5 * gt_widths
    gt_ctr_y = gt_rois[:, 1] + 0.5 * gt_heights

    targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths
    targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights
    targets_dw = np.log(gt_widths / ex_widths)
    targets_dh = np.log(gt_heights / ex_heights)

    targets = np.zeros((ex_rois.shape[0], 4), dtype=np.float32)
    targets[:, 0] = targets_dx
    targets[:, 1] = targets_dy
    targets[:, 2] = targets_dw
    targets[:, 3] = targets_dh
    return targets 

def _compute_targets(ex_rois, gt_rois):
    """Compute bounding-box regression targets for an image. The targets are scale invariance"""

    ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + cfg.EPS
    ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + cfg.EPS
    ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths
    ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights

    gt_widths = gt_rois[:, 2] - gt_rois[:, 0] + cfg.EPS
    gt_heights = gt_rois[:, 3] - gt_rois[:, 1] + cfg.EPS
    gt_ctr_x = gt_rois[:, 0] + 0.5 * gt_widths
    gt_ctr_y = gt_rois[:, 1] + 0.5 * gt_heights

    targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths
    targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights
    targets_dw = np.log(gt_widths / ex_widths)
    targets_dh = np.log(gt_heights / ex_heights)

    targets = np.zeros((ex_rois.shape[0], 4), dtype=np.float32)
    targets[:, 0] = targets_dx
    targets[:, 1] = targets_dy
    targets[:, 2] = targets_dw
    targets[:, 3] = targets_dh
    return targets 

def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str_)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str_)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str_)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def add_bbox_regression_targets(roidb):
    """Add information needed to train bounding-box regressors."""
    assert len(roidb) > 0
    assert 'info_boxes' in roidb[0], 'Did you call prepare_roidb first?'

    num_images = len(roidb)
    # Infer number of classes from the number of columns in gt_overlaps
    num_classes = roidb[0]['gt_overlaps'].shape[1]

    # Compute values needed for means and stds
    # var(x) = E(x^2) - E(x)^2
    class_counts = np.zeros((num_classes, 1)) + cfg.EPS
    sums = np.zeros((num_classes, 4))
    squared_sums = np.zeros((num_classes, 4))
    for im_i in xrange(num_images):
        targets = roidb[im_i]['info_boxes']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 12] == cls)[0]
            if cls_inds.size > 0:
                class_counts[cls] += cls_inds.size
                sums[cls, :] += targets[cls_inds, 14:].sum(axis=0)
                squared_sums[cls, :] += (targets[cls_inds, 14:] ** 2).sum(axis=0)

    means = sums / class_counts
    stds = np.sqrt(squared_sums / class_counts - means ** 2)

    # Normalize targets
    for im_i in xrange(num_images):
        targets = roidb[im_i]['info_boxes']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 12] == cls)[0]
            roidb[im_i]['info_boxes'][cls_inds, 14:] -= means[cls, :]
            if stds[cls, 0] != 0:
                roidb[im_i]['info_boxes'][cls_inds, 14:] /= stds[cls, :]

    # These values will be needed for making predictions
    # (the predicts will need to be unnormalized and uncentered)
    return means.ravel(), stds.ravel() 

def setup(self, bottom, top):
        self._anchors = generate_anchors(cfg.TRAIN.RPN_BASE_SIZE, cfg.TRAIN.RPN_ASPECTS, cfg.TRAIN.RPN_SCALES)
        self._num_anchors = self._anchors.shape[0]

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        layer_params = yaml.load(self.param_str_)
        self._feat_stride = layer_params['feat_stride']

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str_)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def add_bbox_regression_targets(roidb):
    """Add information needed to train bounding-box regressors."""
    assert len(roidb) > 0
    assert 'info_boxes' in roidb[0], 'Did you call prepare_roidb first?'

    num_images = len(roidb)
    # Infer number of classes from the number of columns in gt_overlaps
    num_classes = roidb[0]['gt_overlaps'].shape[1]

    # Compute values needed for means and stds
    # var(x) = E(x^2) - E(x)^2
    class_counts = np.zeros((num_classes, 1)) + cfg.EPS
    sums = np.zeros((num_classes, 4))
    squared_sums = np.zeros((num_classes, 4))
    for im_i in xrange(num_images):
        targets = roidb[im_i]['info_boxes']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 12] == cls)[0]
            if cls_inds.size > 0:
                class_counts[cls] += cls_inds.size
                sums[cls, :] += targets[cls_inds, 14:].sum(axis=0)
                squared_sums[cls, :] += (targets[cls_inds, 14:] ** 2).sum(axis=0)

    means = sums / class_counts
    stds = np.sqrt(squared_sums / class_counts - means ** 2)

    # Normalize targets
    for im_i in xrange(num_images):
        targets = roidb[im_i]['info_boxes']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 12] == cls)[0]
            roidb[im_i]['info_boxes'][cls_inds, 14:] -= means[cls, :]
            if stds[cls, 0] != 0:
                roidb[im_i]['info_boxes'][cls_inds, 14:] /= stds[cls, :]

    # These values will be needed for making predictions
    # (the predicts will need to be unnormalized and uncentered)
    return means.ravel(), stds.ravel() 

def setup(self, bottom, top):
        self._anchors = generate_anchors(cfg.TRAIN.RPN_BASE_SIZE, cfg.TRAIN.RPN_ASPECTS, cfg.TRAIN.RPN_SCALES)
        self._num_anchors = self._anchors.shape[0]

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        layer_params = yaml.load(self.param_str_)
        self._feat_stride = layer_params['feat_stride']

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        layer_params = yaml.load(self.param_str_)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchors = generate_anchors(scales=np.array(anchor_scales))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        try:
            layer_params = yaml.load(self.param_str_)
        except AttributeError:
            layer_params = yaml.load(self.param_str)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchor_ratios = layer_params.get('ratios',(0.5, 1, 2))
        base_size = layer_params.get('base_size', 16)
        self._anchors = generate_anchors(scales=np.array(anchor_scales), base_size=base_size,
                                         ratios=np.array(self._anchor_ratios))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(1, 1, A * height, width)
        # bbox_targets
        top[1].reshape(1, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(1, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(1, A * 4, height, width) 

def setup(self, bottom, top):
        try:
            layer_params = yaml.load(self.param_str_)
        except AttributeError:
            layer_params = yaml.load(self.param_str)
        anchor_scales = layer_params.get('scales', (8, 16, 32))
        self._anchor_ratios = layer_params.get('ratios', (0.5, 1, 2))
        base_size = layer_params.get('base_size', 16)
        self._anchors = generate_anchors(scales=np.array(anchor_scales), base_size=base_size,
                                         ratios=np.array(self._anchor_ratios))
        self._num_anchors = self._anchors.shape[0]
        self._feat_stride = layer_params['feat_stride']

        if DEBUG:
            print 'anchors:'
            print self._anchors
            print 'anchor shapes:'
            print np.hstack((
                self._anchors[:, 2::4] - self._anchors[:, 0::4],
                self._anchors[:, 3::4] - self._anchors[:, 1::4],
            ))
            self._counts = cfg.EPS
            self._sums = np.zeros((1, 4))
            self._squared_sums = np.zeros((1, 4))
            self._fg_sum = 0
            self._bg_sum = 0
            self._count = 0

        # allow boxes to sit over the edge by a small amount
        self._allowed_border = layer_params.get('allowed_border', 0)

        height, width = bottom[0].data.shape[-2:]
        if DEBUG:
            print 'AnchorTargetLayer: height', height, 'width', width

        A = self._num_anchors
        # labels
        top[0].reshape(cfg.TRAIN.IMS_PER_BATCH, 1, A * height, width)
        # bbox_targets
        top[1].reshape(cfg.TRAIN.IMS_PER_BATCH, A * 4, height, width)
        # bbox_inside_weights
        top[2].reshape(cfg.TRAIN.IMS_PER_BATCH, A * 4, height, width)
        # bbox_outside_weights
        top[3].reshape(cfg.TRAIN.IMS_PER_BATCH, A * 4, height, width) 

def add_bbox_regression_targets(roidb):
    """Add information needed to train bounding-box regressors."""
    assert len(roidb) > 0
    assert 'max_classes' in roidb[0], 'Did you call prepare_roidb first?'

    num_images = len(roidb)
    # Infer number of classes from the number of columns in gt_overlaps
    num_classes = roidb[0]['gt_overlaps'].shape[1]
    for im_i in xrange(num_images):
        rois = roidb[im_i]['boxes']
        max_overlaps = roidb[im_i]['max_overlaps']
        max_classes = roidb[im_i]['max_classes']
        roidb[im_i]['bbox_targets'] = \
                _compute_targets(rois, max_overlaps, max_classes, num_classes)

    # Compute values needed for means and stds
    # var(x) = E(x^2) - E(x)^2
    class_counts = np.zeros((num_classes, 1)) + cfg.EPS
    sums = np.zeros((num_classes, 4))
    squared_sums = np.zeros((num_classes, 4))
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 0] == cls)[0]
            if cls_inds.size > 0:
                class_counts[cls] += cls_inds.size
                sums[cls, :] += targets[cls_inds, 1:].sum(axis=0)
                squared_sums[cls, :] += (targets[cls_inds, 1:] ** 2).sum(axis=0)

    means = sums / class_counts
    stds = np.sqrt(squared_sums / class_counts - means ** 2)

    # Normalize targets
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 0] == cls)[0]
            roidb[im_i]['bbox_targets'][cls_inds, 1:] -= means[cls, :]
            if stds[cls, 0] != 0:
                roidb[im_i]['bbox_targets'][cls_inds, 1:] /= stds[cls, :]

    # These values will be needed for making predictions
    # (the predicts will need to be unnormalized and uncentered)
    return means.ravel(), stds.ravel() 

def _compute_targets(rois, overlaps, labels, num_classes):
    """Compute bounding-box regression targets for an image."""
    # Ensure ROIs are floats
    rois = rois.astype(np.float, copy=False)

    # Indices of ground-truth ROIs
    gt_inds = np.where(overlaps == 1)[0]
    # Indices of examples for which we try to make predictions
    ex_inds = []
    for i in xrange(1, num_classes):
        ex_inds.extend( np.where((labels == i) & (overlaps >= cfg.TRAIN.BBOX_THRESH[i-1]))[0] )

    # Get IoU overlap between each ex ROI and gt ROI
    ex_gt_overlaps = utils.cython_bbox.bbox_overlaps(rois[ex_inds, :],
                                                     rois[gt_inds, :])

    # Find which gt ROI each ex ROI has max overlap with:
    # this will be the ex ROI's gt target
    if ex_gt_overlaps.shape[0] != 0:
        gt_assignment = ex_gt_overlaps.argmax(axis=1)
    else:
        gt_assignment = []
    gt_rois = rois[gt_inds[gt_assignment], :]
    ex_rois = rois[ex_inds, :]

    ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + cfg.EPS
    ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + cfg.EPS
    ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths
    ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights

    gt_widths = gt_rois[:, 2] - gt_rois[:, 0] + cfg.EPS
    gt_heights = gt_rois[:, 3] - gt_rois[:, 1] + cfg.EPS
    gt_ctr_x = gt_rois[:, 0] + 0.5 * gt_widths
    gt_ctr_y = gt_rois[:, 1] + 0.5 * gt_heights

    targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths
    targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights
    targets_dw = np.log(gt_widths / ex_widths)
    targets_dh = np.log(gt_heights / ex_heights)

    targets = np.zeros((rois.shape[0], 5), dtype=np.float32)
    targets[ex_inds, 0] = labels[ex_inds]
    targets[ex_inds, 1] = targets_dx
    targets[ex_inds, 2] = targets_dy
    targets[ex_inds, 3] = targets_dw
    targets[ex_inds, 4] = targets_dh
    return targets 

def add_bbox_regression_targets(roidb):
    """Add information needed to train bounding-box regressors."""
    assert len(roidb) > 0
    assert 'max_classes' in roidb[0], 'Did you call prepare_roidb first?'

    num_images = len(roidb)
    # Infer number of classes from the number of columns in gt_overlaps
    num_classes = roidb[0]['gt_overlaps'].shape[1]
    for im_i in xrange(num_images):
        rois = roidb[im_i]['boxes']
        max_overlaps = roidb[im_i]['max_overlaps']
        max_classes = roidb[im_i]['max_classes']
        roidb[im_i]['bbox_targets'] = \
                _compute_targets(rois, max_overlaps, max_classes, num_classes)

    # Compute values needed for means and stds
    # var(x) = E(x^2) - E(x)^2
    class_counts = np.zeros((num_classes, 1)) + cfg.EPS
    sums = np.zeros((num_classes, 4))
    squared_sums = np.zeros((num_classes, 4))
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 0] == cls)[0]
            if cls_inds.size > 0:
                class_counts[cls] += cls_inds.size
                sums[cls, :] += targets[cls_inds, 1:].sum(axis=0)
                squared_sums[cls, :] += (targets[cls_inds, 1:] ** 2).sum(axis=0)

    means = sums / class_counts
    stds = np.sqrt(squared_sums / class_counts - means ** 2)

    # Normalize targets
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 0] == cls)[0]
            roidb[im_i]['bbox_targets'][cls_inds, 1:] -= means[cls, :]
            if stds[cls, 0] != 0:
                roidb[im_i]['bbox_targets'][cls_inds, 1:] /= stds[cls, :]

    # These values will be needed for making predictions
    # (the predicts will need to be unnormalized and uncentered)
    return means.ravel(), stds.ravel() 

def _compute_targets(rois, overlaps, labels, num_classes):
    """Compute bounding-box regression targets for an image."""
    # Ensure ROIs are floats
    rois = rois.astype(np.float, copy=False)

    # Indices of ground-truth ROIs
    gt_inds = np.where(overlaps == 1)[0]
    # Indices of examples for which we try to make predictions
    ex_inds = []
    for i in xrange(1, num_classes):
        ex_inds.extend( np.where((labels == i) & (overlaps >= cfg.TRAIN.BBOX_THRESH[i-1]))[0] )

    # Get IoU overlap between each ex ROI and gt ROI
    ex_gt_overlaps = utils.cython_bbox.bbox_overlaps(rois[ex_inds, :],
                                                     rois[gt_inds, :])

    # Find which gt ROI each ex ROI has max overlap with:
    # this will be the ex ROI's gt target
    if ex_gt_overlaps.shape[0] != 0:
        gt_assignment = ex_gt_overlaps.argmax(axis=1)
    else:
        gt_assignment = []
    gt_rois = rois[gt_inds[gt_assignment], :]
    ex_rois = rois[ex_inds, :]

    ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + cfg.EPS
    ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + cfg.EPS
    ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths
    ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights

    gt_widths = gt_rois[:, 2] - gt_rois[:, 0] + cfg.EPS
    gt_heights = gt_rois[:, 3] - gt_rois[:, 1] + cfg.EPS
    gt_ctr_x = gt_rois[:, 0] + 0.5 * gt_widths
    gt_ctr_y = gt_rois[:, 1] + 0.5 * gt_heights

    targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths
    targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights
    targets_dw = np.log(gt_widths / ex_widths)
    targets_dh = np.log(gt_heights / ex_heights)

    targets = np.zeros((rois.shape[0], 5), dtype=np.float32)
    targets[ex_inds, 0] = labels[ex_inds]
    targets[ex_inds, 1] = targets_dx
    targets[ex_inds, 2] = targets_dy
    targets[ex_inds, 3] = targets_dw
    targets[ex_inds, 4] = targets_dh
    return targets 

def _compute_targets(rois, overlaps, labels, num_classes):
    """Compute bounding-box regression targets for an image."""
    # Ensure ROIs are floats
    rois = rois.astype(np.float, copy=False)

    # Indices of ground-truth ROIs
    gt_inds = np.where(overlaps == 1)[0]
    # Indices of examples for which we try to make predictions
    ex_inds = []
    for i in xrange(1, num_classes):
        ex_inds.extend( np.where((labels == i) & (overlaps >= cfg.TRAIN.BBOX_THRESH))[0] )

    # Get IoU overlap between each ex ROI and gt ROI
    ex_gt_overlaps = utils.cython_bbox.bbox_overlaps(rois[ex_inds, :],
                                                     rois[gt_inds, :])

    # Find which gt ROI each ex ROI has max overlap with:
    # this will be the ex ROI's gt target
    if ex_gt_overlaps.shape[0] != 0:
        gt_assignment = ex_gt_overlaps.argmax(axis=1)
    else:
        gt_assignment = []
    gt_rois = rois[gt_inds[gt_assignment], :]
    ex_rois = rois[ex_inds, :]

    ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + cfg.EPS
    ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + cfg.EPS
    ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths
    ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights

    gt_widths = gt_rois[:, 2] - gt_rois[:, 0] + cfg.EPS
    gt_heights = gt_rois[:, 3] - gt_rois[:, 1] + cfg.EPS
    gt_ctr_x = gt_rois[:, 0] + 0.5 * gt_widths
    gt_ctr_y = gt_rois[:, 1] + 0.5 * gt_heights

    targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths
    targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights
    targets_dw = np.log(gt_widths / ex_widths)
    targets_dh = np.log(gt_heights / ex_heights)

    targets = np.zeros((rois.shape[0], 5), dtype=np.float32)
    targets[ex_inds, 0] = labels[ex_inds]
    targets[ex_inds, 1] = targets_dx
    targets[ex_inds, 2] = targets_dy
    targets[ex_inds, 3] = targets_dw
    targets[ex_inds, 4] = targets_dh
    return targets 

def add_bbox_regression_targets(roidb):
    """Add information needed to train bounding-box regressors."""
    assert len(roidb) > 0
    assert 'max_classes' in roidb[0], 'Did you call prepare_roidb first?'

    num_images = len(roidb)
    # Infer number of classes from the number of columns in gt_overlaps
    num_classes = roidb[0]['gt_overlaps'].shape[1]
    for im_i in xrange(num_images):
        rois = roidb[im_i]['boxes']
        max_overlaps = roidb[im_i]['max_overlaps']
        max_classes = roidb[im_i]['max_classes']
        roidb[im_i]['bbox_targets'] = \
                _compute_targets(rois, max_overlaps, max_classes, num_classes)

    # Compute values needed for means and stds
    # var(x) = E(x^2) - E(x)^2
    class_counts = np.zeros((num_classes, 1)) + cfg.EPS
    sums = np.zeros((num_classes, 4))
    squared_sums = np.zeros((num_classes, 4))
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 0] == cls)[0]
            if cls_inds.size > 0:
                class_counts[cls] += cls_inds.size
                sums[cls, :] += targets[cls_inds, 1:].sum(axis=0)
                squared_sums[cls, :] += (targets[cls_inds, 1:] ** 2).sum(axis=0)

    means = sums / class_counts
    stds = np.sqrt(squared_sums / class_counts - means ** 2)

    # Normalize targets
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        for cls in xrange(1, num_classes):
            cls_inds = np.where(targets[:, 0] == cls)[0]
            roidb[im_i]['bbox_targets'][cls_inds, 1:] -= means[cls, :]
            if stds[cls, 0] != 0:
                roidb[im_i]['bbox_targets'][cls_inds, 1:] /= stds[cls, :]

    # These values will be needed for making predictions
    # (the predicts will need to be unnormalized and uncentered)
    return means.ravel(), stds.ravel() 

def compute_bbox_target_normalization(roidb):
    num_images = len(roidb)
    # Infer number of classes from the number of columns in gt_overlaps
    num_classes = roidb[0]['gt_overlaps'].shape[1]
    for im_i in xrange(num_images):
        rois = roidb[im_i]['boxes']
        max_overlaps = roidb[im_i]['max_overlaps']
        max_classes = roidb[im_i]['max_classes']
        roidb[im_i]['bbox_targets'], roidb[im_i]['gt_ind_assignments'] = \
                _compute_targets(rois, max_overlaps, max_classes)

    class_counts = np.zeros((num_classes, 1)) + cfg.EPS
    sums = np.zeros((num_classes, 4))
    squared_sums = np.zeros((num_classes, 4))
    for im_i in xrange(num_images):
        targets = roidb[im_i]['bbox_targets']
        image_target_classes = np.unique(targets[:, 0]).astype(int)
        for cls in image_target_classes:
            if(cls > 0):
                cls_inds = np.where(targets[:, 0] == cls)[0]
                class_counts[cls] += cls_inds.size
                sums[cls, :] += targets[cls_inds, 1:].sum(axis=0)
                squared_sums[cls, :] += \
                        (targets[cls_inds, 1:] ** 2).sum(axis=0)

    means = sums / class_counts
    stds = np.sqrt(squared_sums / class_counts - means ** 2)
    np.save(cfg.TRAIN.BBOX_TARGET_NORMALIZATION_FILE, {'means': means, 'stds': stds})

    if cfg.TRAIN.BBOX_NORMALIZE_TARGETS:
        print "Normalizing targets"
        for im_i in xrange(num_images):
            targets = roidb[im_i]['bbox_targets']
            image_target_classes = np.unique(targets[:, 0]).astype(int)
            for cls in image_target_classes:
                if(cls > 0):
                    cls_inds = np.where(targets[:, 0] == cls)[0]
                    roidb[im_i]['bbox_targets'][cls_inds, 1:] -= means[cls, :]
                    roidb[im_i]['bbox_targets'][cls_inds, 1:] /= stds[cls, :]
    else:
        print "NOT normalizing targets"

    return means, stds