Play generated audio as it is generating. (#457)

k2-fsa · Dec 2, 2023 · 99ff6a8 · 99ff6a8
1 parent 539b27e
commit 99ff6a8
Show file tree

Hide file tree

Showing 20 changed files with 876 additions and 79 deletions.
diff --git a/cmake/cmake_extension.py b/cmake/cmake_extension.py
@@ -143,6 +143,7 @@ def build_extension(self, ext: setuptools.extension.Extension):
         binaries += ["sherpa-onnx-vad-microphone"]
         binaries += ["sherpa-onnx-vad-microphone-offline-asr"]
         binaries += ["sherpa-onnx-offline-tts"]
+        binaries += ["sherpa-onnx-offline-tts-play"]
 
         if is_windows():
             binaries += ["kaldi-native-fbank-core.dll"]

diff --git a/cmake/espeak-ng-for-piper.cmake b/cmake/espeak-ng-for-piper.cmake
@@ -5,6 +5,7 @@ function(download_espeak_ng_for_piper)
   set(espeak_ng_URL2 "")
   set(espeak_ng_HASH "SHA256=8a48251e6926133dd91fcf6cb210c7c2e290a9b578d269446e2d32d710b0dfa0")
 
+  set(BUILD_ESPEAK_NG_TESTS OFF CACHE BOOL "" FORCE)
   set(USE_ASYNC OFF CACHE BOOL "" FORCE)
   set(USE_MBROLA OFF CACHE BOOL "" FORCE)
   set(USE_LIBSONIC OFF CACHE BOOL "" FORCE)
@@ -106,10 +107,12 @@ function(download_espeak_ng_for_piper)
   if(SHERPA_ONNX_ENABLE_PYTHON AND WIN32)
     install(TARGETS
       espeak-ng
+      ucd
     DESTINATION ..)
   else()
     install(TARGETS
       espeak-ng
+      ucd
     DESTINATION lib)
   endif()
 
@@ -120,6 +123,7 @@ function(download_espeak_ng_for_piper)
   if(WIN32 AND BUILD_SHARED_LIBS)
     install(TARGETS
       espeak-ng
+      ucd
     DESTINATION bin)
   endif()
 endfunction()

diff --git a/mfc-examples/NonStreamingSpeechRecognition/sherpa-onnx-deps.props b/mfc-examples/NonStreamingSpeechRecognition/sherpa-onnx-deps.props
@@ -14,6 +14,9 @@
         sherpa-onnx-fst.lib;
         kaldi-native-fbank-core.lib;
         onnxruntime.lib;
+        piper_phonemize.lib;
+        espeak-ng.lib;
+        ucd.lib;
     </SherpaOnnxLibraries>
   </PropertyGroup>
   <ItemDefinitionGroup>

diff --git a/mfc-examples/NonStreamingTextToSpeech/sherpa-onnx-deps.props b/mfc-examples/NonStreamingTextToSpeech/sherpa-onnx-deps.props
@@ -14,6 +14,9 @@
         sherpa-onnx-fst.lib;
         kaldi-native-fbank-core.lib;
         onnxruntime.lib;
+        piper_phonemize.lib;
+        espeak-ng.lib;
+        ucd.lib;
     </SherpaOnnxLibraries>
   </PropertyGroup>
   <ItemDefinitionGroup>

diff --git a/mfc-examples/StreamingSpeechRecognition/sherpa-onnx-deps.props b/mfc-examples/StreamingSpeechRecognition/sherpa-onnx-deps.props
@@ -14,6 +14,9 @@
         sherpa-onnx-fst.lib;
         kaldi-native-fbank-core.lib;
         onnxruntime.lib;
+        piper_phonemize.lib;
+        espeak-ng.lib;
+        ucd.lib;
     </SherpaOnnxLibraries>
   </PropertyGroup>
   <ItemDefinitionGroup>