neuralfilter 0.1.9

neural attention filter

Neural Filter

Functions

neuralfilter.generate(x, force=False)

• x: Array with resolution $(H, W, C)$
  The dimension $C$ is recommended as 1.
• force: If you want to force this operation when the dimension $C$ is higher than 1, set 'force' as True.

neuralfilter.batch_generate(x, force=False)

• The batch processing mode of 'generate'.
• x: Array with resolution $(N, H, W, C)$
  The dimension $C$ is recommended as 1.
• force: If you want to force this operation when the dimension $C$ is higher than 1, set 'force' as True.

batch_generate(x, force=False)

Usage

import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
import neuralfilter

(x_tr, y_tr), (x_te, y_te) = tf.keras.datasets.mnist.load_data()

idx = 1
a = neuralfilter.generate(np.expand_dims(x_tr[idx], axis=-1))

plt.figure(figsize=(9, 3), dpi=100)
plt.subplot(1, 3, 1)
plt.imshow(x_tr[idx], cmap='gray')
plt.xticks([], [])
plt.yticks([], [])

plt.subplot(1, 3, 2)
plt.imshow(a, cmap='jet')
plt.xticks([], [])
plt.yticks([], [])

plt.subplot(1, 3, 3)
plt.imshow(x_tr[idx], cmap='gray')
plt.imshow(a, cmap='jet', alpha=0.5)
plt.xticks([], [])
plt.yticks([], [])

plt.tight_layout()
plt.savefig('sample.png')
plt.show()