Helper function for creating a resnet_v1 bottleneck block. Args: scope: The scope of the block. base_depth: The depth of the bottleneck laye

Helper function for creating a resnet bottleneck block. Args: scope: The scope of the block. base_depth: The depth of the bottleneck layer f

Generator for v1 ResNet models. This function generates a family of ResNet v1 models. See the resnet_v1_*() methods for specific model instantiati

Pads the input along the spatial dimensions independently of input size. Pads the input such that if it was used in a convolution with 'VALID' paddi

Uniformly sample from the given range. Args val_range: A pair of lower and upper bound.

Generator for v2 (preactivation) ResNet models. This function generates a family of ResNet v2 models. See the resnet_v2_*() methods for specific m

Returns `slim.arg_scope` with all None arguments removed. Arguments: funcs: Functions to pass to `arg_scope`. **kwargs: Arguments to pass to

Clip and convert an image to np.uint8. Args image: Image to clip.

Sets arg scope defaults for all items present in defaults. Args: defaults: dictionary/list of pairs, containing a mapping from function to a

Evenly splits num, num_ways so each piece is a multiple of divisible_by.

Computes area of boxes. Arguments: boxes: a float tensor with shape [N, 4]. Returns: a float tensor with shape [N] representi

Returns: [anchors]: each anchors is a SxSx NUM_ANCHOR_RATIOS x4 array.

Creates a split convolution. Split convolution splits the input and output into 'num_blocks' blocks of approximately the same size each, and onl

Defines Mobilenet training scope. Usage: with tf.contrib.slim.arg_scope(mobilenet.training_scope()): logits, endpoints = mobilenet_v2.mo

Pads the input along the spatial dimensions independently of input size. Pads the input such that if it was used in a convolution with 'VALID' paddi

Determines if the graph has any variables. Args: sess: TensorFlow Session. Returns: Bool.

Data augmentation function.

Parses input tensorflow graph into GraphDef proto.

Parses input tensorflow graph into MetaGraphDef proto.

Parses input tensorflow Saver into SaverDef proto.

Adjust brightness of an image Args image: Image to adjust. delta: Brightness offset between -1 and 1 added to the pixel values.

Adjust contrast of an image. Args image: Image to adjust. factor: A factor for adjusting contrast.

Adjust hue of an image. Args image: Image to adjust. delta: An interval between -1 and 1 for the amount added to the hue channel

Adjust saturation of an image. Args image: Image to adjust. factor: An interval for the factor multiplying the saturation values

Calculate the average gradient for each shared variable across all towers. Note that this function provides a synchronization point across al

Decodes a batch of box encodings with respect to the anchors. Arguments: box_encodings: a float tensor with shape [batch_size, num_anchor

Arguments: boxes: a float tensor with shape [batch_size, N, 4]. scores: a float tensor with shape [batch_size, N]. score_

:param boxes: [n,x,y,x,y] :return: [4n,x,y]

Decode relative codes to boxes. Arguments: codes: a float tensor with shape [N, 4], anchor-encoded boxes of the format [tx1, t

Encode boxes with respect to anchors. Arguments: boxes: a float tensor with shape [N, 4]. anchors: a float tensor with shape [N, 4

r"""Compute focal loss for predictions. Multi-labels Focal loss formula: FL = -alpha * (z-p)^gamma * log(p) -(1-alpha) * p^gamma *

Converts all variables in a graph and checkpoint into constants. Args: input_graph: A `GraphDef` file to load. input_saver: A TensorFlow Sav

Converts all variables in a graph and checkpoint into constants. Args: input_graph_def: A `GraphDef`. input_saver_def: A `SaverDef` (optiona

Get all anchors in the largest possible image, shifted, floatbox Args: max_size(int) : h w stride (int): the stride of anchor

Get all anchors in the largest possible image, shifted, floatbox Args: stride (int): the stride of anchors. s

Returns: [anchors]: each anchors is a SxSx NUM_ANCHOR_RATIOS x4 array.

Applies avg pool to produce 1x1 output. NOTE: This function is funcitonally equivalenet to reduce_mean, but it has baked in average pool which has

Compute pairwise intersection areas between boxes. Arguments: boxes1: a float tensor with shape [N, 4]. boxes2: a float tensor with sha

Compute the jaccard overlap of two sets of boxes. The jaccard overlap is simply the intersection over union of two boxes. Args:

Compute the jaccard overlap of two sets of boxes. The jaccard overlap is simply the intersection over union of two boxes. Args:

A usual L1 smooth loss. Arguments: predictions: a float tensor with shape [batch_size, num_anchors, 4], representing the (enc

Creates mobilenet V2 network. Inference mode is created by default. To create training use training_scope below. with tf.contrib.slim.arg_scope(

Mobilenet base network. Constructs a network from inputs to the given final endpoint. By default the network is constructed in inference mode. To

:return: [fname,lbel] type:list

Postprocess outputs of the network. Returns: boxes: a float tensor with shape [batch_size, N, 4]. scores: a float ten