alexander | f42f568 | 2021-07-16 11:30:56 +0100 | [diff] [blame] | 1 | # Copyright © 2021 Arm Ltd and Contributors. All rights reserved. |
| 2 | # SPDX-License-Identifier: MIT |
| 3 | |
| 4 | """Utilities for speech recognition apps.""" |
| 5 | |
| 6 | import numpy as np |
| 7 | |
| 8 | |
| 9 | def decode(model_output: np.ndarray, labels: dict) -> list: |
| 10 | """Decodes the integer encoded results from inference into a string. |
| 11 | |
| 12 | Args: |
| 13 | model_output: Results from running inference. |
| 14 | labels: Dictionary of labels keyed on the classification index. |
| 15 | |
| 16 | Returns: |
| 17 | Decoded string. |
| 18 | """ |
| 19 | results = [labels[np.argmax(model_output)], model_output[0][0][np.argmax(model_output)]] |
| 20 | |
| 21 | return results |
| 22 | |
| 23 | |
| 24 | def display_text(text: list): |
| 25 | """Presents the results on the console. |
| 26 | |
| 27 | Args: |
| 28 | text: Results of performing ASR on the input audio data. |
| 29 | """ |
| 30 | print('Classification: %s' % text[0]) |
| 31 | print('Probability: %s' % text[1]) |