From 7e9696f15635a83ff810cc86b79c0932cb4c8a27 Mon Sep 17 00:00:00 2001
From: Djraemon <154407769+Djraemon@users.noreply.github.com>
Date: Mon, 4 Nov 2024 11:04:52 +0800
Subject: [PATCH] Fix audio format 2.0 (#2186)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

* 添加了对音频文件扩展名是否为.mp3的补丁，是mp3格式则转化为wav格式

* 增加检测音频文件是否为mp3格式的补丁

* 完善对音频文件后缀名的检查，若文件后缀不是.wav，则转化为wav

* 增加音频文件后缀名检查；音频文件无效时抛出错误

* 在paraformer、vad两个模型中加入对音频文件后缀的检查，并将非wav格式转为wav格式

* 修改wav_path的数据类型，使demo能够顺利运行
---
 .../onnxruntime/demo_paraformer_online.py       |  3 ++-
 .../onnxruntime/funasr_onnx/paraformer_bin.py   | 16 ++++++++++++++++
 .../funasr_onnx/paraformer_online_bin.py        | 17 +++++++++++++++++
 .../python/onnxruntime/funasr_onnx/vad_bin.py   | 17 +++++++++++++++++
 4 files changed, 52 insertions(+), 1 deletion(-)

diff --git a/runtime/python/onnxruntime/demo_paraformer_online.py b/runtime/python/onnxruntime/demo_paraformer_online.py
index bb72b41a1..cca5357ab 100644
--- a/runtime/python/onnxruntime/demo_paraformer_online.py
+++ b/runtime/python/onnxruntime/demo_paraformer_online.py
@@ -3,7 +3,8 @@ from funasr_onnx.paraformer_online_bin import Paraformer
 from pathlib import Path
 
 model_dir = "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online"
-wav_path = ["{}/.cache/modelscope/hub/{}/example/asr_example.wav".format(Path.home(), model_dir)]
+#wav_path = ["{}/.cache/modelscope/hub/{}/example/asr_example.wav".format(Path.home(), model_dir)]
+wav_path = "{}/.cache/modelscope/hub/{}/example/asr_example.wav".format(Path.home(), model_dir)
 
 chunk_size = [5, 10, 5]
 model = Paraformer(
diff --git a/runtime/python/onnxruntime/funasr_onnx/paraformer_bin.py b/runtime/python/onnxruntime/funasr_onnx/paraformer_bin.py
index 4f35fcc86..bde7f442e 100644
--- a/runtime/python/onnxruntime/funasr_onnx/paraformer_bin.py
+++ b/runtime/python/onnxruntime/funasr_onnx/paraformer_bin.py
@@ -175,7 +175,23 @@ class Paraformer:
         plt.savefig(plotname, bbox_inches="tight")
 
     def load_data(self, wav_content: Union[str, np.ndarray, List[str]], fs: int = None) -> List:
+        def convert_to_wav(input_path, output_path):
+            from pydub import AudioSegment
+            try:
+                audio = AudioSegment.from_mp3(input_path)
+                audio.export(output_path, format="wav")
+                print("音频文件为mp3格式，已转换为wav格式")
+                
+            except Exception as e:
+                print(f"转换失败:{e}")
+
         def load_wav(path: str) -> np.ndarray:
+            if not path.lower().endswith('.wav'):
+                import os
+                input_path = path
+                path = os.path.splitext(path)[0]+'.wav'
+                convert_to_wav(input_path,path) #将mp3格式转换成wav格式
+
             waveform, _ = librosa.load(path, sr=fs)
             return waveform
 
diff --git a/runtime/python/onnxruntime/funasr_onnx/paraformer_online_bin.py b/runtime/python/onnxruntime/funasr_onnx/paraformer_online_bin.py
index ddd857d37..3f63ea020 100644
--- a/runtime/python/onnxruntime/funasr_onnx/paraformer_online_bin.py
+++ b/runtime/python/onnxruntime/funasr_onnx/paraformer_online_bin.py
@@ -208,7 +208,24 @@ class Paraformer:
         return asr_res
 
     def load_data(self, wav_content: Union[str, np.ndarray, List[str]], fs: int = None) -> List:
+        
+        def convert_to_wav(input_path, output_path):
+            from pydub import AudioSegment
+            try:
+                audio = AudioSegment.from_mp3(input_path)
+                audio.export(output_path, format="wav")
+                print("音频文件为mp3格式，已转换为wav格式")
+                
+            except Exception as e:
+                print(f"转换失败:{e}")
+
         def load_wav(path: str) -> np.ndarray:
+            if not path.lower().endswith('.wav'):
+                import os
+                input_path = path
+                path = os.path.splitext(path)[0]+'.wav'
+                convert_to_wav(input_path,path) #将mp3格式转换成wav格式
+
             waveform, _ = librosa.load(path, sr=fs)
             return waveform
 
diff --git a/runtime/python/onnxruntime/funasr_onnx/vad_bin.py b/runtime/python/onnxruntime/funasr_onnx/vad_bin.py
index 92928a840..3f21004bf 100644
--- a/runtime/python/onnxruntime/funasr_onnx/vad_bin.py
+++ b/runtime/python/onnxruntime/funasr_onnx/vad_bin.py
@@ -143,7 +143,24 @@ class Fsmn_vad:
         return segments
 
     def load_data(self, wav_content: Union[str, np.ndarray, List[str]], fs: int = None) -> List:
+        
+        def convert_to_wav(input_path, output_path):
+            from pydub import AudioSegment
+            try:
+                audio = AudioSegment.from_mp3(input_path)
+                audio.export(output_path, format="wav")
+                print("音频文件为mp3格式，已转换为wav格式")
+                
+            except Exception as e:
+                print(f"转换失败:{e}")
+
         def load_wav(path: str) -> np.ndarray:
+            if not path.lower().endswith('.wav'):
+                import os
+                input_path = path
+                path = os.path.splitext(path)[0]+'.wav'
+                convert_to_wav(input_path,path) #将mp3格式转换成wav格式
+
             waveform, _ = librosa.load(path, sr=fs)
             return waveform