Rotate masks counter-clockwise by 90 degrees. Args: masks: rank 3 float32 tensor with shape [num_instances, height, width] representing i

Evaluates metrics defined in evaluators and returns summaries. This function loads the latest checkpoint in checkpoint_dirs and evaluates all met

Samples a mini-batch of proposals to be sent to the box classifier. Helper function for self._postprocess_rpn. Args: proposal_boxlist:

Scale logits then apply softmax.

Performs random crop. Note: boxes will be clipped to the crop. Keypoint coordinates that are outside the crop will be set to NaN, which is consis

Subsample a collection of selected indices to a desired neg:pos ratio. This function takes a subset of M indices (indexing into a large anchor

Creates tensorflow summaries for the input boxes and anchors. This function creates four summaries corresponding to the average number (over

Unpads proposals and applies hard mining. Args: proposal_boxlists: A list of `batch_size` BoxLists each representing `self.max_num_

Unpads proposals and samples a minibatch for second stage. Args: proposal_boxes: A float tensor with shape [batch_size, num_proposa

Updates `configs` to reflect the new training batch size. The configs dictionary is updated in place, and hence not returned. Args: configs:

Updates the classification/localization weight loss ratio. Detection models usually define a loss weight for both classification and objectness.

Updates `configs` to reflect new number of eval steps per evaluation.

Updates the alpha value for a sigmoid focal loss. The configs dictionary is updated in place, and hence not returned. Args: configs: Diction

Updates the gamma value for a sigmoid focal loss. The configs dictionary is updated in place, and hence not returned. Args: configs: Diction

Update a pipeline configuration parameter based on a generic key/value. Args: configs: Dictionary of pipeline configuration protos. key: A

Update grouth truth statitistics. 1. Difficult boxes are ignored when counting the number of ground truth instances as done in Pascal VOC dev

Updates grouth truth statistics. Args: groundtruth_class_tuples: A numpy array of structures shape [M, 1], representing the clas

Updates `configs` to reflect the new initial learning rate. This function updates the initial learning rate. For learning rate schedules, all oth

Updates the label map path for both train and eval input readers. The configs dictionary is updated in place, and hence not returned. Args:

Updates the mask type for both train and eval input readers. The configs dictionary is updated in place, and hence not returned. Args: confi

Updates `configs` to reflect the new momentum value. Momentum is only supported for RMSPropOptimizer and MomentumOptimizer. For any other optimiz

Adds a new constant learning rate.

Adds a new cosine decay learning rate.

Adds a new exponential decay learning rate.

Adds a learning rate schedule.

Updates `configs` to reflect new number of training steps.

Updates the eval config option to use or not use moving averages. The configs dictionary is updated in place, and hence not returned. Args:

Checks if a label map is valid. Args: label_map: StringIntLabelMap to validate. Raises: ValueError: if label map is invalid.

Computes area of boxes. Args: boxlist: BoxList holding N boxes Returns: a numpy array with shape [N*1] representing box areas

From current list of trackers and new detections, output matched detections, unmatchted trackers, unmatched detections.

Performs box voting as described in S. Gidaris and N. Komodakis, ICCV 2015. Performs box voting as described in 'Object detection via a multi-regio

Builds hard example miner based on the config. Args: config: A losses_pb2.HardExampleMiner object. classification_weight: Classification lo

Clears the state to prepare for a fresh evaluation.

Clears detections.

Clips the input tensor along the first dimension up to the length. Args: t: the input tensor, assuming the rank is at least 1. length: a te

Concatenate list of BoxLists. This op concatenates a list of input BoxLists into a larger BoxList. It also handles concatenation of BoxList fiel

Concatenate list of BoxLists. This op concatenates a list of input BoxLists into a larger BoxList. It also handles concatenation of BoxList fiel

Convert the KITTI detection dataset to TFRecords. Args: data_dir: The full path to the unzipped folder containing the unzipped data from

Loads label map proto and returns categories list compatible with eval. This function loads a label map and returns a list of dicts, each of which

Creates dictionary of COCO compatible categories keyed by category id. Args: categories: a list of dicts, each of which has the following keys:

Creates a configs dictionary from pipeline_pb2.TrainEvalPipelineConfig. Args: pipeline_config: pipeline_pb2.TrainEvalPipelineConfig proto objec

Creates `Estimator`, input functions, and steps. Args: run_config: A `RunConfig`. hparams: A `HParams`. pipeline_config_path: A path to

Sets up reader, prefetcher and returns input queue. Args: batch_size_per_clone: batch size to use per clone. create_tensor_dict_fn: functio

Creates a pipeline_pb2.TrainEvalPipelineConfig from configs dictionary. This function performs the inverse operation of create_configs_from_pipel

Creates a predict `input` function for `Estimator`. Args: model_config: A model_pb2.DetectionModel. Returns: `input_fn` for `Estimator`

Converts image and annotations to a tf.Example proto. Args: image: dict with keys: [u'license', u'file_name', u'coco_url', u'height', u'w

Creates a train `input` function for `Estimator`. Args: train_config: A train_pb2.TrainConfig. train_input_config: An input_reader_pb2.Inpu

Convert XML derived dict to tf.Example proto. Notice that this function normalizes the bounding box coordinates provided by the raw data. Args

Convert XML derived dict to tf.Example proto. Notice that this function normalizes the bounding box coordinates provided by the raw data. Args

Adds a bounding box to an image (numpy array). Bounding box coordinates can be specified in either absolute (pixel) or normalized coordinates by

Adds a bounding box to an image (numpy array). Bounding box coordinates can be specified in either absolute (pixel) or normalized coordinates by

Draws bounding boxes on image. Args: image: a PIL.Image object. boxes: a 2 dimensional numpy array of [N, 4]: (ymin, xmin, ymax, xmax).

Draws bounding boxes on image. Args: image: a PIL.Image object. boxes: a 2 dimensional numpy array of [N, 4]: (ymin, xmin, ymax, xmax).

Draws keypoints on an image. Args: image: a PIL.Image object. keypoints: a numpy array with shape [num_keypoints, 2]. color: color to d

Draws keypoints on an image. Args: image: a PIL.Image object. keypoints: a numpy array with shape [num_keypoints, 2]. color: color to d

Draws keypoints on an image (numpy array). Args: image: a numpy array with shape [height, width, 3]. keypoints: a numpy array with shape [n

Draws keypoints on an image (numpy array). Args: image: a numpy array with shape [height, width, 3]. keypoints: a numpy array with shape [n

Evaluates detections and returns a dictionary of metrics.

Constructs the graph, executes it on TPU and returns the result. Args: graph_fn: a callable that constructs the tensorflow graph to test. T

Extracts features from preprocessed inputs. This function is responsible for extracting feature maps from preprocessed images. Args:

Filters out annotations from the unused classes and dontcare regions. Filters out the annotations that belong to classes we do now wish to use and

Filter to keep only boxes with field entries equal to the given value. Args: boxlist: BoxList holding N boxes. field: field name for filter

Converts all variables in a graph and checkpoint into constants.

Returns all fields.

Gets the class name from a file. Args: file_name: The file name to get the class name from. ie. "american_pit_bull_terrier_105.j

Reads config from a file containing pipeline_pb2.TrainEvalPipelineConfig. Args: pipeline_config_path: Path to pipeline_pb2.TrainEvalPipelineCon

Returns the default mapping from a preprocessor function to its args. Args: include_label_scores: If True, preprocessing functions will modify

Builds a callable to compute depth (output channels) of conv filters. Args: depth_multiplier: a multiplier for the nominal depth. min_depth

Returns the evaluator class according to eval_config, valid for categories. Args: eval_config: evaluation configurations. categories: a lis

Parses config for graph rewriter. Args: graph_rewriter_config_file: file path to the graph rewriter config. Returns: graph_rewriter_pb2.

Dequeues batch and constructs inputs to object detection model. Args: input_queue: BatchQueue object holding enqueued tensor_dicts. num_cla

Returns the learning rate type for training. Args: optimizer_config: An optimizer_pb2.Optimizer. Returns: The type of the learning rate.

Computes height and width of boxes in boxlist. Args: boxlist: BoxList holding N boxes scope: name scope. Returns: Height: A tensor w

Computes pairwise intersection-over-area between box and mask collections. Intersection-over-area (ioa) between two masks mask1 and mask2 is define

Computes pairwise intersection-over-area between box collections. Intersection-over-area (ioa) between two boxes box1 and box2 is defined as thei

Computes pairwise intersection-over-area between box collections. intersection-over-area (IOA) between two boxes box1 and box2 is defined as thei

Computes pairwise intersection-over-union between box collections. Args: boxes1: a numpy array with shape [N, 4] holding N boxes. boxes2: a

Computes pairwise intersection-over-union between box collections. Args: boxlist1: BoxList holding N boxes boxlist2: BoxList holding M boxe