Hugging Face's logo

AXERA-TECH
/
OpenWakeWord.AXERA

keyword-spotting
Model card Files
xet
 Community
OpenWakeWord.AXERA
File size: 2,738 Bytes
759d0dd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
# openWakeWord.AXERA

[openWakeWord](https://github.com/dscripka/openWakeWord) 语音唤醒 AXERA 推理 demo。

## 功能

- 支持 AX650 和 AX630C；
- 支持 Python 和 C++ 板端推理；

## 目录结构

```text
OpenWakeWord.AXERA/
├── models/
│   ├── 650/          # AX650 axmodel
│   └── 630C/         # AX630C axmodel
├── config/           # STFT/mel 权重
├── audio/openwakeword/
├── scripts/          # Python 推理代码
├── cpp/              # C++ 源码、编译和运行脚本
└── README.md
```


## Python

```bash
创建虚拟环境
conda create -n OpenWakeWord python=3.10
conda activate OpenWakeWord

安装 [pyaxengine](https://github.com/AXERA-TECH/pyaxengine/releases/latest)：
pip install axengine-x.x.x-py3-none-any.whl

pip install -r requirements.txt
```

AX650：

```bash
python scripts/openwakeword_ax650.py
```

AX630C：

```bash
python scripts/openwakeword_ax630c.py
```

自定义音频目录、阈值和输出文件：

```bash
python scripts/openwakeword_ax630c.py \
  --audio-dir audio/openwakeword \
  --threshold 0.5 \
  --output outputs/openwakeword_ax630c.json
```

## C++

### 编译环境

AX650 和 AX630C 共用 Arm GNU 9.2 AArch64 交叉编译器：

<https://developer.arm.com/-/media/Files/downloads/gnu-a/9.2-2019.12/binrel/gcc-arm-9.2-2019.12-x86_64-aarch64-none-linux-gnu.tar.xz>

```bash
mkdir -p cpp/toolchains
wget -P cpp/toolchains \
  https://developer.arm.com/-/media/Files/downloads/gnu-a/9.2-2019.12/binrel/gcc-arm-9.2-2019.12-x86_64-aarch64-none-linux-gnu.tar.xz
tar -xf cpp/toolchains/gcc-arm-9.2-2019.12-x86_64-aarch64-none-linux-gnu.tar.xz \
  -C cpp/toolchains
```

BSP SDK 下载地址：

| 平台 | BSP SDK |
| --- | --- |
| AX650 | <https://github.com/ZHEQIUSHUI/assets/releases/download/ax_3.6.2/msp_50_3.10.2.zip> |
| AX630C | <https://github.com/ZHEQIUSHUI/assets/releases/download/ax_3.6.2/msp_20e_3.0.0.zip> |

下载并整理到 `cpp/toolchains/`：

```bash
bash cpp/download_bsp.sh all
```

### 交叉编译

```bash
export TOOLCHAIN_ROOT="cpp/toolchains/gcc-arm-9.2-2019.12-x86_64-aarch64-none-linux-gnu"

bash cpp/build_ax650.sh
bash cpp/build_ax630c.sh
```

编译结果：

```text
cpp/bin/openwakeword_ax650
cpp/bin/openwakeword_ax630c
```

### 板端批量测试

AX650：

```bash
THRESHOLD=0.5 bash cpp/run_batch_openwakeword_ax650.sh
```

AX630C：

```bash
THRESHOLD=0.5 bash cpp/run_batch_openwakeword_ax630c.sh
```

结果分别写入 `outputs/cpp_ax650_batch/` 和 `outputs/cpp_ax630c_batch/`。

详细 C++ 说明见 `cpp/README.md`。

## 参考

- [openWakeWord](https://github.com/dscripka/openWakeWord)
- [Pulsar2](https://pulsar2-docs.readthedocs.io/en/latest/pulsar2/introduction.html)