Constructs a minibatch sampler. Args: positive_fraction: desired fraction of positive examples (scalar in [0,1]) in the batch.

Construct Object Detection Target Assigner. Args: similarity_calc: a RegionSimilarityCalculator matcher: an object_detection.core.Mat

Constructs a batch queue holding tensor_dict. Args: tensor_dict: dictionary of tensors to batch. batch_size: batch size. batch_

Constructor. Args: num_classes: number of classes. Note that num_classes *does not* include background categories that might be impl

Constructor. Args: hierarchy: labels hierarchy as JSON file.

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

InceptionV2 Feature Extractor for SSD Models. Args: is_training: whether the network is in training mode. depth_multiplier: float dep

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

SSD Resnet50 V1 FPN feature extractor based on Resnet v1 architecture. Args: is_training: whether the network is in training mode. de

SSD Resnet101 V1 FPN feature extractor based on Resnet v1 architecture. Args: is_training: whether the network is in training mode. d

SSD Resnet152 V1 FPN feature extractor based on Resnet v1 architecture. Args: is_training: whether the network is in training mode. d

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

MobileNetV1 Feature Extractor for SSD Models. Args: is_training: whether the network is in training mode. depth_multiplier: float dep

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

MobileNetV2 Feature Extractor for SSD Models. Mobilenet v2 (experimental), designed by sandler@. More details can be found in //knowledge/cer

InceptionV3 Feature Extractor for SSD Models. Args: is_training: whether the network is in training mode. depth_multiplier: float dep

Constructor. Args: is_training: See base class. first_stage_features_stride: See base class. batch_norm_trainable: See base cla

MobileNetV1 Feature Extractor for Embedded-friendly SSD Models. Args: is_training: whether the network is in training mode. depth_mul

Constructor. Args: is_training: whether the network is in training mode. depth_multiplier: float depth multiplier for feature extract

SSDMetaArch Constructor. TODO(rathodv,jonathanhuang): group NMS parameters + score converter into a class and loss parameters into a class an

RFCNMetaArch Constructor. Args: is_training: A boolean indicating whether the training version of the computation graph should be c

Constructor. Args: is_training: A boolean indicating whether the training version of the computation graph should be constructed.

FasterRCNNMetaArch Constructor. Args: is_training: A boolean indicating whether the training version of the computation graph shoul

Construct ArgMaxMatcher. Args: matched_threshold: Threshold for positive matches. Positive if sim >= matched_threshold, where sim i

Constructs a Matcher. Args: use_matmul_gather: Force constructed match objects to use matrix multiplication based gather instead of

Constructor for FasterRcnnBoxCoder. Args: scale_factors: List of 4 positive scalars to scale ty, tx, th and tw. If set to None, doe

Constructor for SquareBoxCoder. Args: scale_factors: List of 3 positive scalars to scale ty, tx, and tl. If set to None, does not p

Constructor for KeypointBoxCoder. Args: num_keypoints: Number of keypoints to encode/decode. scale_factors: List of 4 positive scalar

Constructor for MeanStddevBoxCoder. Args: stddev: The standard deviation used to encode and decode boxes.

Compute pairwise IOU similarity between the two BoxLists. Args: boxlist1: BoxList holding N boxes. boxlist2: BoxList holding M boxes.

Compute matrix of (negated) sq distances. Args: boxlist1: BoxList holding N boxes. boxlist2: BoxList holding M boxes. Returns:

Compute pairwise IOA similarity between the two BoxLists. Args: boxlist1: BoxList holding N boxes. boxlist2: BoxList holding M boxes.

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, code_size] representing the (enco

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, code_size] representing the (enco

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, 4] representing the decoded predi

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, num_classes] representing the pre

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, num_classes] representing the pre

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, num_classes] representing the pre

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, num_classes] representing the pre

Compute loss function. Args: prediction_tensor: A float tensor of shape [batch_size, num_anchors, num_classes] representing the pre

Creates loss function for a DetectionModel. Args: input_queue: BatchQueue object holding enqueued tensor_dicts. create_model_fn: A function

Decode relative codes to boxes. Args: rel_codes: a tensor representing N anchor-encoded boxes. anchors: BoxList of anchors. Retu

Decodes relative codes to boxes. Args: rel_codes: a tensor representing N anchor-encoded boxes. anchors: BoxList of anchors. Ret

Decode relative codes to boxes and keypoints. Args: rel_codes: a tensor with shape [N, 4 + 2 * num_keypoints] representing N anchor

Decode. Args: rel_codes: a tensor representing N anchor-encoded boxes. anchors: BoxList of anchors. Returns: boxes: BoxLis

Decode PNG instance segmentation masks and stack into dense tensor. The instance segmentation masks are reshaped to [num_instances, height, w

Encode a box collection with respect to anchor collection. Args: boxes: BoxList holding N boxes to be encoded. anchors: BoxList of an

Encodes a box collection with respect to an anchor collection. Args: boxes: BoxList holding N boxes to be encoded. anchors: BoxList o

Encode a box and keypoint collection with respect to anchor collection. Args: boxes: BoxList holding N boxes and keypoints to be encoded. B

Encode a box collection with respect to anchor collection. Args: boxes: BoxList holding N boxes to be encoded. anchors: BoxList of N

Returns input that accepts a batch of PNG or JPEG strings. Returns: a tuple of input placeholder and the output decoded images.

Returns `features` and `labels` tensor dictionaries for evaluation. Args: params: Parameter dictionary passed from the estimator. Retu

Extracts second stage box classifier features. This function reconstructs the "second half" of the NASNet-A network after the part defined in

Extracts second stage box classifier features. Args: proposal_feature_maps: A 4-D float tensor with shape [batch_size * self.max_nu

Extracts second stage box classifier features. This function reconstructs the "second half" of the PNASNet network after the part defined in

Extracts second stage box classifier features. Args: proposal_feature_maps: A 4-D float tensor with shape [batch_size * self.max_nu

Extracts second stage box classifier features. Args: proposal_feature_maps: A 4-D float tensor with shape [batch_size * self.max_nu

Extracts second stage box classifier features. This function reconstructs the "second half" of the Inception ResNet v2 network after the part

Extracts first stage RPN features. Extracts features using the first half of the NASNet network. We construct the network in `align_feature_m

Extracts first stage RPN features. Args: preprocessed_inputs: A [batch, height, width, channels] float32 tensor representing a batc

Extracts first stage RPN features. Extracts features using the first half of the PNASNet network. We construct the network in `align_feature_

Extracts first stage RPN features. Args: preprocessed_inputs: A [batch, height, width, channels] float32 tensor representing a batc

Extracts first stage RPN features. Args: preprocessed_inputs: A [batch, height, width, channels] float32 tensor representing a batc

Extracts first stage RPN features. Extracts features using the first half of the Inception Resnet v2 network. We construct the network in `al

Generates a collection of bounding boxes to be used as anchors. Args: feature_map_shape_list: list of pairs of convnet layer resolutions in

Generates a collection of bounding boxes to be used as anchors. Currently we require the input image shape to be statically defined. That is

Generates a collection of bounding boxes to be used as anchors. The number of anchors generated for a single grid with shape MxM where we pla

Returns input placeholder and a 4-D uint8 image tensor.

Tries to match each column of the similarity matrix to a row. Args: similarity_matrix: tensor of shape [N, M] representing any similarity

Bipartite matches a collection rows and columns. A greedy bi-partite. TODO(rathodv): Add num_valid_columns options to match only that many column

Performs matching when the rows of similarity matrix are empty. When the rows are empty, all detections are false positives. So we return

Performs matching when the rows of similarity matrix are non empty. Returns: matches: int32 tensor indicating the row each column matc

Randomly samples anchors for one image. Args: inputs: a list of 2 inputs. First one is a tensor of shape [num_anchors, num_classes]

Computes encoded object locations and corresponding confidences. Args: image_features: A list of float tensors of shape [batch_size, height

Optionally computes encoded object locations, confidences, and masks. Flattens image_features and applies fully connected ops (with no non-li