def __init__(self, params):
super(ShapeMaskModel, self).__init__(params)
self._params = params
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._retinanet_head_fn = factory.retinanet_head_generator(params)
self._shape_prior_head_fn = factory.shapeprior_head_generator(params)
self._coarse_mask_fn = factory.coarsemask_head_generator(params)
self._fine_mask_fn = factory.finemask_head_generator(params)
self._outer_box_scale = params.architecture.outer_box_scale
# Loss function.
self._cls_loss_fn = losses.RetinanetClassLoss(
params.retinanet_loss, params.architecture.num_classes)
self._box_loss_fn = losses.RetinanetBoxLoss(params.retinanet_loss)
self._box_loss_weight = params.retinanet_loss.box_loss_weight
# Mask loss function.
self._shapemask_prior_loss_fn = losses.ShapemaskMseLoss()
self._shapemask_loss_fn = losses.ShapemaskLoss()
self._shape_prior_loss_weight = (
params.shapemask_loss.shape_prior_loss_weight)
self._coarse_mask_loss_weight = (
params.shapemask_loss.coarse_mask_loss_weight)
self._fine_mask_loss_weight = (
params.shapemask_loss.fine_mask_loss_weight)
# Predict function.
self._generate_detections_fn = postprocess_ops.MultilevelDetectionGenerator(
params.architecture.min_level, params.architecture.max_level,
params.postprocess)
def __init__(self, params):
super(MaskrcnnModel, self).__init__(params)
self._anchor_params = params.anchor
self._include_mask = params.architecture.include_mask
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._rpn_head_fn = factory.rpn_head_generator(params.rpn_head)
self._generate_rois_fn = roi_ops.ROIGenerator(params.roi_proposal)
self._sample_rois_fn = sampling_ops.ROISampler(params.roi_sampling)
self._sample_masks_fn = sampling_ops.MaskSampler(params.mask_sampling)
self._frcnn_head_fn = factory.fast_rcnn_head_generator(
params.frcnn_head)
if self._include_mask:
self._mrcnn_head_fn = factory.mask_rcnn_head_generator(
params.mrcnn_head)
# Loss function.
self._rpn_score_loss_fn = losses.RpnScoreLoss(params.rpn_score_loss)
self._rpn_box_loss_fn = losses.RpnBoxLoss(params.rpn_box_loss)
self._frcnn_class_loss_fn = losses.FastrcnnClassLoss()
self._frcnn_box_loss_fn = losses.FastrcnnBoxLoss(params.frcnn_box_loss)
if self._include_mask:
self._mask_loss_fn = losses.MaskrcnnLoss()
self._generate_detections_fn = postprocess_ops.GenericDetectionGenerator(
params.postprocess)
self._transpose_input = params.train.transpose_input
def __init__(self, params):
super(RetinanetModel, self).__init__(params)
# For eval metrics.
self._params = params
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._head_fn = factory.retinanet_head_generator(params.retinanet_head)
# Loss function.
self._cls_loss_fn = losses.RetinanetClassLoss(params.retinanet_loss)
self._box_loss_fn = losses.RetinanetBoxLoss(params.retinanet_loss)
self._box_loss_weight = params.retinanet_loss.box_loss_weight
self._keras_model = None
# Predict function.
self._generate_detections_fn = postprocess_ops.MultilevelDetectionGenerator(
params.postprocess)
self._transpose_input = params.train.transpose_input
assert not self._transpose_input, 'Transpose input is not supportted.'
# Input layer.
input_shape = (params.retinanet_parser.output_size +
[params.retinanet_parser.num_channels])
self._input_layer = tf.keras.layers.Input(
shape=input_shape,
name='',
batch_size=self._params.train.batch_size,
dtype=tf.bfloat16 if self._use_bfloat16 else tf.float32)
def __init__(self, params):
super(AttributeMaskrcnnModel, self).__init__(params)
self._params = params
self._include_mask = params.architecture.include_mask
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._rpn_head_fn = factory.rpn_head_generator(params)
self._generate_rois_fn = roi_ops.ROIGenerator(params.roi_proposal)
self._sample_rois_fn = roi_sampler.ROISampler(params.roi_sampling)
self._sample_masks_fn = target_ops.MaskSampler(
params.architecture.mask_target_size,
params.mask_sampling.num_mask_samples_per_image)
self._frcnn_head_fn = factory.fast_rcnn_head_generator(params)
if self._include_mask:
self._mrcnn_head_fn = factory.mask_rcnn_head_generator(params)
# Loss function.
self._rpn_score_loss_fn = losses.RpnScoreLoss(params.rpn_score_loss)
self._rpn_box_loss_fn = losses.RpnBoxLoss(params.rpn_box_loss)
self._frcnn_class_loss_fn = losses.FastrcnnClassLoss()
self._frcnn_attribute_loss_fn = attribute_loss.FastrcnnAttributeLoss()
self._frcnn_box_loss_fn = losses.FastrcnnBoxLoss(params.frcnn_box_loss)
if self._include_mask:
self._mask_loss_fn = losses.MaskrcnnLoss()
self._generate_detections_fn = postprocess_ops.GenericDetectionGenerator(
params.postprocess)
def __init__(self, params):
super(SegmentationModel, self).__init__(params)
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._head_fn = factory.segmentation_head_generator(
params.segmentation_head)
self._num_classes = params.segmentation_head.num_classes
# Loss function.
self._loss_fn = losses.SegmentationLoss(params.segmentation_loss)
self._l2_weight_decay = params.train.l2_weight_decay
self._transpose_input = params.train.transpose_input
def __init__(self, params):
super(SegmentationModel, self).__init__(params)
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._head_fn = factory.segmentation_head_generator(params)
self._num_classes = params.architecture.num_classes
self._level = params.segmentation_head.level
# Loss function.
self._loss_fn = losses.SegmentationLoss(params.segmentation_loss)
self._use_aspp = params.architecture.use_aspp
self._use_pyramid_fusion = params.architecture.use_pyramid_fusion
def __init__(self, params):
super(ViLDModel, self).__init__(params)
self._params = params
self._include_mask = params.architecture.include_mask
self._losses = params.train.losses
# feature distill
self._feat_distill = params.architecture.visual_feature_distill
if self._feat_distill == 'None':
self._feat_distill = None
self._feat_distill_dim = params.architecture.visual_feature_dim
self._max_distill_rois = params.architecture.max_num_rois
self._feat_distill_weight = params.architecture.feat_distill_weight
self._normalize_feat_during_training = params.architecture.normalize_feat_during_training
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._rpn_head_fn = factory.rpn_head_generator(params)
self._generate_rois_fn = roi_ops.ROIGenerator(params.roi_proposal)
self._sample_rois_fn = target_ops.ROISampler(params.roi_sampling)
self._sample_masks_fn = target_ops.MaskSampler(
params.architecture.mask_target_size,
params.mask_sampling.num_mask_samples_per_image)
self._frcnn_head_fn = factory.vild_fast_rcnn_head_generator(params)
if self._include_mask:
self._mrcnn_head_fn = factory.mask_rcnn_head_generator(params)
# Loss function.
self._rpn_score_loss_fn = losses.RpnScoreLoss(params.rpn_score_loss)
self._rpn_box_loss_fn = losses.RpnBoxLoss(params.rpn_box_loss)
self._frcnn_class_loss_fn = vild_losses.FastrcnnClassLoss(
params.frcnn_class_loss)
self._frcnn_box_loss_fn = losses.FastrcnnBoxLoss(
params.frcnn_box_loss,
class_agnostic_bbox_pred=params.frcnn_head.class_agnostic_bbox_pred
)
if self._include_mask:
self._mask_loss_fn = losses.MaskrcnnLoss()
self._generate_detections_fn = postprocess_ops.GenericDetectionGenerator(
params.postprocess,
discard_background=params.postprocess.discard_background,
visual_feature_distill=self._feat_distill)
def __init__(self, params):
super(CascadeMaskrcnnModel, self).__init__(params)
self._params = params
self._include_mask = params.architecture.include_mask
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._rpn_head_fn = factory.rpn_head_generator(params)
self._generate_rois_fn = roi_ops.ROIGenerator(params.roi_proposal)
self._sample_rois_fn = target_ops.ROISampler(params.roi_sampling)
self._sample_masks_fn = target_ops.MaskSampler(
params.architecture.mask_target_size,
params.mask_sampling.num_mask_samples_per_image)
self._frcnn_head_fn = factory.fast_rcnn_head_generator(params)
if self._include_mask:
self._mrcnn_head_fn = factory.mask_rcnn_head_generator(params)
# Loss function.
self._rpn_score_loss_fn = losses.RpnScoreLoss(params.rpn_score_loss)
self._rpn_box_loss_fn = losses.RpnBoxLoss(params.rpn_box_loss)
self._frcnn_class_loss_fn = losses.FastrcnnClassLoss()
self._frcnn_box_loss_fn = losses.FastrcnnBoxLoss(
params.frcnn_box_loss, params.frcnn_head.class_agnostic_bbox_pred)
if self._include_mask:
self._mask_loss_fn = losses.MaskrcnnLoss()
# IoU thresholds for additional FRCNN heads in Cascade mode. 'fg_iou_thresh'
# is the first threshold.
self._cascade_iou_thresholds = params.roi_sampling.cascade_iou_thresholds
self._num_roi_samples = params.roi_sampling.num_samples_per_image
# Weights for the regression losses for each FRCNN layer.
# TODO(golnazg): makes this param configurable.
self._cascade_layer_to_weights = [
[10.0, 10.0, 5.0, 5.0],
[20.0, 20.0, 10.0, 10.0],
[30.0, 30.0, 15.0, 15.0],
]
self._class_agnostic_bbox_pred = params.frcnn_head.class_agnostic_bbox_pred
self._cascade_class_ensemble = params.frcnn_head.cascade_class_ensemble
self._generate_detections_fn = postprocess_ops.GenericDetectionGenerator(
params.postprocess)
def __init__(self, params):
super(RetinanetModel, self).__init__(params)
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._head_fn = factory.retinanet_head_generator(params.retinanet_head)
# Loss function.
self._cls_loss_fn = losses.RetinanetClassLoss(params.retinanet_loss)
self._box_loss_fn = losses.RetinanetBoxLoss(params.retinanet_loss)
self._box_loss_weight = params.retinanet_loss.box_loss_weight
# Predict function.
self._generate_detections_fn = postprocess_ops.MultilevelDetectionGenerator(
params.postprocess)
self._transpose_input = params.train.transpose_input
def __init__(self, params):
super(RetinanetModel, self).__init__(params)
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._head_fn = factory.retinanet_head_generator(params.retinanet_head)
# Loss function.
self._cls_loss_fn = losses.RetinanetClassLoss(params.retinanet_loss)
self._box_loss_fn = losses.RetinanetBoxLoss(params.retinanet_loss)
self._box_loss_weight = params.retinanet_loss.box_loss_weight
# Predict function.
self._generate_detections_fn = postprocess.GenerateOneStageDetections(
params.postprocess)
self._l2_weight_decay = params.architecture.l2_weight_decay
def __init__(self, params):
super(RetinanetModel, self).__init__(params)
self._params = params
# Architecture generators.
self._backbone_fn = factory.backbone_generator(params)
self._fpn_fn = factory.multilevel_features_generator(params)
self._head_fn = factory.retinanet_head_generator(params)
# Loss function.
self._cls_loss_fn = losses.RetinanetClassLoss(
params.retinanet_loss, params.architecture.num_classes)
self._box_loss_fn = losses.RetinanetBoxLoss(params.retinanet_loss)
self._box_loss_weight = params.retinanet_loss.box_loss_weight
self._focal_loss_normalizer_momentum = (
params.retinanet_loss.normalizer_momentum)
# Predict function.
self._generate_detections_fn = postprocess_ops.MultilevelDetectionGenerator(
params.architecture.min_level, params.architecture.max_level,
params.postprocess)