@ -2,6 +2,7 @@
*Author: Jael Gu*
< br / >
## Desription
@ -11,34 +12,36 @@ This operator is built on top of [VGGish](https://github.com/tensorflow/models/t
The model is a [VGG ](https://arxiv.org/abs/1409.1556 ) variant pre-trained with a large scale of audio dataset [AudioSet ](https://research.google.com/audioset ).
As suggested, it is suitable to extract features at high level or warm up a larger model.
< br / >
## Code Example
Generate embeddings for the audio "test.wav".
*Write the pipeline in simplified style*
*Write the pipeline in simplified style*:
```python
from towhee import dc
import towhee
dc.glob('test.wav')
.audio_decode()
.time_window(range=10)
.audio_embedding.vggish()
.show()
towhee.glob('test.wav') \
.audio_decode() \
.time_window(range=10) \
.audio_embedding.vggish() \
.show()
```
| [-0.4931737, -0.40068552, -0.032327592, ...] shape=(10, 128) |
*Write a same pipeline with explicit inputs/outputs name specifications:*
```python
from towhee import dc
dc.glob['path']('test.wav')
.audio_decode['path', 'audio']()
.time_window['audio', 'frames'](range=10)
.audio_embedding.vggish['frames', 'vecs']()
.select('vecs')
.to_vec()
import towhee
towhee.glob['path']('test.wav') \
.audio_decode['path', 'audio']() \
.time_window['audio', 'frames'](range=10) \
.audio_embedding.vggish['frames', 'vecs']() \
.select('vecs') \
.to_vec()
```
[array([[-0.4931737 , -0.40068552, -0.03232759, ..., -0.33428153,
0.1333081 , -0.25221825],
@ -54,6 +57,8 @@ dc.glob['path']('test.wav')
[-0.4886143 , -0.40098593, -0.03175077, ..., -0.3325425 ,
0.13271847, -0.25159872]], dtype=float32)]
< br / >
## Factory Constructor
Create the operator via the following factory method
@ -62,34 +67,33 @@ Create the operator via the following factory method
**Parameters:**
*weights_path: str*
*weights_path: str*
The path to model weights. If None, it will load default model weights.
The path to model weights. If None, it will load default model weights.
*framework: str*
*framework: str*
The framework of model implementation.
The framework of model implementation.
Default value is "pytorch" since the model is implemented in Pytorch.
## Interface
< br / >
An audio embedding operator generates vectors in numpy.ndarray given an audio file path or a [towhee audio ](link/to/AudioFrame/api/doc ).
## Interface
An audio embedding operator generates vectors in numpy.ndarray given an audio file path or towhee audio frames.
**Parameters:**
*Union[str, towhee.types.Audio]*
*Union[str, towhee.types.Audio (a sub-class of numpy.ndarray) ]*
The audio path or link in string.
The audio path or link in string.
Or audio input data in towhee audio frames.
The input data should represent for an audio longer than 0.9s.
**Returns**:
*numpy.ndarray*
*numpy.ndarray*
Audio embeddings in shape (num_clips, 128).
Audio embeddings in shape (num_clips, 128).
Each embedding stands for features of an audio clip with length of 0.9s.
