tf_softmax_layer.py

'''
softmax_layer.py : generic SoftMax inference() method for TensorFlow

Adapted from:

    https://github.com/darksigma/Fundamentals-of-Deep-Learning-Book/blob/master/fdl_examples/chapter3/logistic_regression_updated.py

Copyright (C) 2017 Jack Baird, Alex Cantrell, Keith Denning, Rajwol Joshi, 
Simon D. Levy, Will McMurtry, Jacob Rosen

This file is part of AirSimTensorFlow

MIT License
'''

import tensorflow as tf

def inference(x, xsize, ysize, W_vals=0, b_vals=0):
    '''
    This is a general-purpose softmax inference layer implementation.
    '''
    W_init = tf.constant_initializer(value=W_vals)
    b_init = tf.constant_initializer(value=b_vals)
    W = tf.get_variable('W', [xsize, ysize], initializer=W_init)
    b = tf.get_variable('b', [ysize],        initializer=b_init)
    output = tf.nn.softmax(tf.matmul(x, W) + b)

    return output