源码编译 FunASR for windows on arm

源码编译 FunASR for windows on arm

这里有编译好的，直接下载使用

https://github.com/turingevo/FunASR-build/releases

编译

• 1 下载 onnxruntime-win-arm64：
  https://github.com/microsoft/onnxruntime/releases/download/v1.16.1/onnxruntime-win-arm64-1.16.1.zip
  解压到
  C:\workspace\program\onnxruntime-win-arm64-1.16.1

• 2 下载 ffmpeg for windows on arm
  https://github.com/wmx-github/ffmpeg-wos-arm64-build/releases/tag/v6.1

• 3 下载 openssl for windows on arm
  https://github.com/turingevo/openssl-build

• 4 打开cmake-gui
  选中配置为 Visual Studio 17 2022
  打开源码路径 C:/workspace/FunASR/runtime/websocket
  添加变量：

FFMPEG_DIR=C:\workspace\program\ffmpeg-wos-arm64
ONNXRUNTIME_DIR=C:\workspace\program\onnxruntime-win-arm64-1.16.1
OPENSSL_ROOT_DIR=C:\workspace\program\OpenSSL

configure 、generate 用vsual studio 打开，编译

生成文件在这里：C:\workspace\FunASR\runtime\websocket\build\bin\Release

使用

1 下载模型：

在modelscope ： https://www.modelscope.cn/models
直接搜索模型名称，然后下载，放在路径下C:\workspace\models\

iic/speech_fsmn_vad_zh-cn-16k-common-onnx 
iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch  
iic/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-onnx
iic/punc_ct-transformer_cn-en-common-vocab471067-large-onnx 
iic/speech_ngram_lm_zh-cn-ai-wesp-fst 
thuduj12/fst_itn_zh 

2 下载测试音频：

https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav

3 启用server:

在FunASR/runtime/ 路径下，新建 run_server.bat 文件：

@echo off
:: Start the command directly
"C:\workspace\FunASR\runtime\websocket\build\bin\Release\funasr-wss-server.exe" ^
    --download-model-dir "C:\workspace\models" ^
    --model-dir "C:\workspace\models\speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-onnx"
    --itn-dir "C:\workspace\models\fst_itn_zh" ^
    --lm-dir "C:\workspace\models\speech_ngram_lm_zh-cn-ai-wesp-fst" ^
    --decoder-thread-num 4 ^
    --model-thread-num 4 ^
    --io-thread-num 4 ^
    --port 10095 ^
    --certfile "C:\workspace\FunASR\runtime\ssl_key\server.crt" ^
    --keyfile "C:\workspace\FunASR\runtime\ssl_key\server.key" ^
	--hotword "C:\workspace\FunASR\runtime\websocket\hotwords.txt"

4 运行client:

在FunASR/runtime/ 路径下打开cmd,输入：

websocket/build/bin/funasr-wss-client --server-ip 127.0.0.1 --port 10095 --wav-path asr_example_zh.wav

输出：

{"is_final":false,"mode":"offline","stamp_sents":[{"end":5195,"punc":"。","start":880,"text_seg":"欢 迎 大 家 来 体 验 达 摩 院 推 出 的 语 音 识 别 模 型","ts_list":[[880,1120],[1120,1380],[1380,1540],[1540,1780],[1780,2020],[2020,2180],[2180,2480],[2480,2600],[2600,2780],[2780,3040],[3040,3240],[3240,3480],[3480,3699],[3699,3900],[3900,4180],[4180,4420],[4420,4620],[4620,4780],[4780,5195]]}],"text":"欢迎大家来体验达摩院推出的语音识别模型。","timestamp":"[[880,1120],[1120,1380],[1380,1540],[1540,1780],[1780,2020],[2020,2180],[2180,2480],[2480,2600],[2600,2780],[2780,3040],[3040,3240],[3240,3480],[3480,3699],[3699,3900],[3900,4180],[4180,4420],[4420,4620],[4620,4780],[4780,5195]]","wav_name":"wav_default_id"}

提取其中的text字段的值，就是解码结果
"欢迎大家来体验达摩院推出的语音识别模型。"