Added utility functions to generate multilabel azimuth encoding from azimuth...

Added utility functions to generate multilabel azimuth encoding from azimuth values, and corresponding test.

core/utils.py

@@ -231,3 +231,32 @@ def azimuth_to_left_center_right_onehot(labels, in_place=False):
     labels[(labels > 0) & (labels < 180)] = 2  # right side
     labels = tf.one_hot(labels, depth=3)
     return labels
+
+
+def is_power_of_two(n):
+    # Code from https://stackoverflow.com/users/6045800/tomerikoo answer,
+    # see: https://stackoverflow.com/questions/57025836/how-to-check-if-a-given-number-is-a-power-of-two
+    return (n & (n - 1) == 0) and n != 0
+
+
+def generate_multilabel_from_azimuth(azimuth, n_labels):
+    """
+    n_labels must be a power of 2, i.e. it should exist an integer k such that n_labels=2^k
+    Take azimuth labels, and encode in a binary word of length 2+2^2+2^3+...+2^k
+    :param n_labels:
+    :return:
+    """
+    if not is_power_of_two(n_labels):
+        raise ValueError('n_labels must be a power of two')
+
+    import math
+    import tensorflow as tf
+    n_iters = int(math.log2(n_labels))
+    labels = []
+    for i in range(n_iters):
+        bins = np.linspace(start=0, stop=360, num=2 ** (i + 1), endpoint=False)
+        inds = np.digitize(x=azimuth, bins=bins)
+        one_hot = tf.one_hot(inds - 1, depth=2 ** (i + 1))
+        labels.append(one_hot)
+    labels = tf.concat(values=labels, axis=1)
+    return labels.shape[-1], labels

pytests/test_utils.py

@@ -20,3 +20,46 @@ def test_get_gtf_kernels():
     from core.utils import get_gtf_kernels
     kernels = get_gtf_kernels("../gtf/")
     assert kernels.shape == (512, 20)
+
+
+def test_generate_multilabel_from_azimuth():
+    from core.utils import generate_multilabel_from_azimuth
+    import numpy as np
+    azimuth = np.array([0, 10, 45, 80, 90, 95, 170, 190, 355, 360])
+    n_labels = 32
+    total_labels, labels = generate_multilabel_from_azimuth(azimuth=azimuth, n_labels=n_labels)
+    expected_total_labels = 2 + 4 + 8 + 16 + 32
+    expected_labels = np.array(
+        [[1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
+         [0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1],
+         [0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+          0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1]])
+
+    assert total_labels == expected_total_labels
+    assert np.array_equal(a1=labels, a2=expected_labels)