first commit

TTS.py

@@ -0,0 +1,24 @@
+import torch
+import torchaudio
+from speechbrain.inference.TTS import Tacotron2
+from speechbrain.inference.vocoders import HIFIGAN
+from speechbrain.inference.TTS import MSTacotron2
+#%%
+
+def TTS(INPUT_TEXT: object,CHOİCE:object) -> object:
+    ms_tacotron2 = MSTacotron2.from_hparams(source="speechbrain/tts-mstacotron2-libritts", savedir="pretrained_models/tts-mstacotron2-libritts")
+    hifi_gan = HIFIGAN.from_hparams(source="speechbrain/tts-hifigan-libritts-22050Hz", savedir="pretrained_models/tts-hifigan-libritts-22050Hz")
+
+    if CHOİCE == "Female":
+        tacotron2 = Tacotron2.from_hparams(source="speechbrain/tts-tacotron2-ljspeech", savedir="tmpdir_tts")
+        hifi_gan = HIFIGAN.from_hparams(source="speechbrain/tts-hifigan-ljspeech", savedir="tmpdir_vocoder")
+        mel_output, mel_length, alignment = tacotron2.encode_text(INPUT_TEXT)
+        waveforms = hifi_gan.decode_batch(mel_output)
+        torchaudio.save('Output/base-TTS.wav',waveforms.squeeze(1), 22050)
+    elif CHOİCE == "Male":
+        REFERENCE_SPEECH = "Voice Samples/natural_m.wav"
+        mel_outputs, mel_lengths, alignments = ms_tacotron2.clone_voice(INPUT_TEXT, REFERENCE_SPEECH)
+        waveforms = hifi_gan.decode_batch(mel_outputs)
+        torchaudio.save("Output/base-TTS.wav", waveforms[0], 22050)
+
+

app.py

@@ -0,0 +1,36 @@
+import gradio as gr
+from TTS import TTS
+import transformers
+
+def text_to_speech(text, choice):
+    TTS(text, choice)
+    return "Output/base-TTS.wav"
+
+def convert_to_speech(text, choice):
+    if text:
+        output_file = text_to_speech(text, choice=choice)
+        with open(output_file, 'rb') as audio_file:
+            audio_bytes = audio_file.read()
+        return (audio_bytes, "Conversion successful!")
+    else:
+        return (None, "Please enter some text to convert.")
+
+def app(text, choice):
+    audio, message = convert_to_speech(text, choice)
+    return audio, message
+
+iface = gr.Interface(
+    fn=app,
+    inputs=[
+        gr.Textbox(lines=2, placeholder="Enter text here...", label="Text Input"),
+        gr.Radio(choices=["Female", "Male"], label="Speaker")
+    ],
+    outputs=[
+        gr.Audio(type="filepath", label="Output Audio"),
+        gr.Textbox(label="Message")
+    ],
+    title="Stars AI Text to Speech Conversion App",
+    description="Convert text to speech with a female or male voice."
+)
+
+iface.launch(share=True)

requirements.txt

@@ -0,0 +1,7 @@
+streamlit==1.34.0
+noisereduce==3.0.2
+torch
+torchaudio
+speechbrain
+scipy
+transformers