....

.gitignore

@@ -1,3 +1,5 @@
 # DotEnv configuration
 .env
+#
+*.pyc
 

tabs/chatbot_tab.py

@@ -209,27 +209,42 @@ def run():
         if audio_bytes:
             st.write("**"+tr("Vendeur")+" :**\n")
             # Fonction pour générer et jouer le texte en speech
-            st.audio(audio_bytes, format="audio/wav")
+            st.audio(audio_bytes, format="audio/wav", autoplay=False)
             try:
-                # Create a BytesIO object from the audio stream
-                audio_stream_bytesio = io.BytesIO(audio_bytes)
+                detection = False
+                if detection:
+                    # Create a BytesIO object from the audio stream
+                    audio_stream_bytesio = io.BytesIO(audio_bytes)
 
-                # Read the WAV stream using wavio
-                wav = wavio.read(audio_stream_bytesio) 
+                    # Read the WAV stream using wavio
+                    wav = wavio.read(audio_stream_bytesio) 
 
-                # Extract the audio data from the wavio.Wav object
-                audio_data = wav.data
+                    # Extract the audio data from the wavio.Wav object
+                    audio_data = wav.data
 
-                # Convert the audio data to a NumPy array
-                audio_input = np.array(audio_data, dtype=np.float32)
-                audio_input = np.mean(audio_input, axis=1)/32768
-                        
-                result = model_speech.transcribe(audio_input)
-                Lang_detected = result["language"]
-                query = result["text"]
-                st.write(tr("Langue détectée")+" : "+Lang_detected)
+                    # Convert the audio data to a NumPy array
+                    audio_input = np.array(audio_data, dtype=np.float32)
+                    audio_input = np.mean(audio_input, axis=1)/32768
+                            
+                    result = model_speech.transcribe(audio_input)
+                    Lang_detected = result["language"]
+                    query = result["text"]
+                    st.write(tr("Langue détectée")+" : "+Lang_detected)
+                
+                else:
+                    # Avec l'aide de la bibliothèque speech_recognition de Google
+                    Lang_detected = "fr"
+                    # Transcription google
+                    audio_stream = sr.AudioData(audio_bytes, 32000, 2) 
+                    r = sr.Recognizer()
+                    query = r.recognize_google(audio_stream, language = Lang_detected)
+                    
+    
+
+                
                 # Transcription Whisper (si result a été préalablement calculé)
                 st.write("**Vendeur :** "+query)
+                st.write("")
             
                 if query != "":
                     input_messages = [HumanMessage(query)]