def prediction_network(x_h, x_m, test=False):
with tf.variable_scope("network") as sc:
if test:
sc.reuse_variables()
with tf.variable_scope("hand"):
flat_h = network_arm(x_h)
with tf.variable_scope("main"):
flat_m = network_arm(x_m)
combined = tf.concat(2, [flat_h, flat_m])
combined, past, future = fork(combined)
encoded = gru_last(combined, 512 * 2, 1, batch_size, "lstm_encoder")
decoded_past = gru_last(combined, 512 * 2, 1, batch_size, "lstm_past")
decoded_future = gru_last(combined, 512 * 2, 1, batch_size,
"lstm_future")
return tf.add(tf.nn.l2_loss(decoded_past, past),
tf.nn.l2_loss(decoded_future, future))
def classification_network(x_h, x_m, y, test=False):
with tf.variable_scope("network") as sc:
if test:
sc.reuse_variables()
with tf.variable_scope("hand"):
flat_h = network_arm(x_h)
with tf.variable_scope("main"):
flat_m = network_arm(x_m)
combined = tf.concat(2, [flat_h, flat_m])
flat = gru_last(combined, 512 * 2, 2, batch_size, "lstm")
with tf.variable_scope("out") as sc:
b_output = bias(20, 0.1)
w_output = weights([1024, 20], 0.02)
output = tf.matmul(flat, w_output) + b_output
return output, tf.reduce_mean(
tf.nn.softmax_cross_entropy_with_logits(output, y))