Update app.py

app.py

@@ -1,153 +1,125 @@
 import streamlit as st
 from TTS.api import TTS
 import torch
-tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2", gpu=True).to("cuda")
-# Dependencies
-%cd /content/
-
 import locale
-locale.getpreferredencoding = lambda: "UTF-8"
-
-!git clone https://github.com/justinjohn0306/Wav2Lip
-!cd Wav2Lip && pip install -r requirements_colab.txt
-
-%cd /content/Wav2Lip
-
-!wget "https://www.adrianbulat.com/downloads/python-fan/s3fd-619a316812.pth" -O "face_detection/detection/sfd/s3fd.pth"
-!wget 'https://github.com/justinjohn0306/Wav2Lip/releases/download/models/wav2lip.pth' -O 'checkpoints/wav2lip.pth'
-!wget 'https://github.com/justinjohn0306/Wav2Lip/releases/download/models/wav2lip_gan.pth' -O 'checkpoints/wav2lip_gan.pth'
-!wget 'https://github.com/justinjohn0306/Wav2Lip/releases/download/models/resnet50.pth' -O 'checkpoints/resnet50.pth'
-!wget 'https://github.com/justinjohn0306/Wav2Lip/releases/download/models/mobilenet.pth' -O 'checkpoints/mobilenet.pth'
-
 import subprocess
 import assemblyai as aai
 import requests
 import moviepy.editor as mp
 
-class translation:
-  def __init__(self,video_path,original_language,target_language):
-    self.video_path= video_path
-    self.original_language = original_language
-    self.target_language = target_language
-
+# Load TTS model
+tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2", gpu=True).to("cuda")
 
+# Function to translate video
+class Translation:
+    def __init__(self, video_path, original_language, target_language):
+        self.video_path = video_path
+        self.original_language = original_language
+        self.target_language = target_language
 
-  def org_language_parameters(self,original_language):
+    def org_language_parameters(self, original_language):
         if original_language == 'English':
-            self.lan_code='en'
-        elif original_language =='German':
-            self.lan_code='de'
-        elif original_language =='French':
-            self.lan_code='fr'
-        elif original_language =='Spanish':
-            self.lan_code='es'
+            self.lan_code = 'en'
+        elif original_language == 'German':
+            self.lan_code = 'de'
+        elif original_language == 'French':
+            self.lan_code = 'fr'
+        elif original_language == 'Spanish':
+            self.lan_code = 'es'
         else:
             self.lan_code = ''
 
-  def target_language_parameters(self,target_language):
+    def target_language_parameters(self, target_language):
         if target_language == 'English':
-            self.tran_code='en'
-        elif target_language =='German':
-            self.tran_code='de'
-        elif target_language =='French':
-            self.tran_code='fr'
-        elif target_language =='Spanish':
-            self.tran_code='es'
+            self.tran_code = 'en'
+        elif target_language == 'German':
+            self.tran_code = 'de'
+        elif target_language == 'French':
+            self.tran_code = 'fr'
+        elif target_language == 'Spanish':
+            self.tran_code = 'es'
         else:
             self.tran_code = ''
 
-  def extract_audio(self):
-          video = mp.VideoFileClip(self.video_path)
-          audio = video.audio
-          audio_path = "output_audio.wav"
-          audio.write_audiofile(audio_path)
-          print("Audio extracted successfully!")
-          return audio_path
-  def transcribe_audio(self,audio_path):
-      aai.settings.api_key = "c29eb650444a4ae4be6a787ebb15d5e2"
-      config = aai.TranscriptionConfig(language_code=self.lan_code)
-      transcriber = aai.Transcriber(config=config)
-      transcript = transcriber.transcribe(audio_path)
-      transcript_text = transcript.text
-      return transcript_text
-      if transcript.status == aai.TranscriptStatus.error:
-          print(transcript.error)
-          return None
-
-
-  def translate_text(self,transcript_text):
-          base_url = "https://api.cognitive.microsofttranslator.com"
-          endpoint = "/translate"
-          headers = {
-              "Ocp-Apim-Subscription-Key": "cd226bb1f3644276bea01d82dd861cbb",
-              "Content-Type": "application/json",
-              "Ocp-Apim-Subscription-Region": "southeastasia"
-          }
-          params = {
-              "api-version": "3.0",
-              "from": self.lan_code,
-              "to": self.tran_code
-          }
-          body = [{"text": transcript_text}]
-          response = requests.post(base_url + endpoint, headers=headers, params=params, json=body)
-          response.raise_for_status()
-          translation = response.json()[0]["translations"][0]["text"]
-          return translation
-
-
-
-
-        #generate audio
-  def generate_audio(self,translated_text):
-      tts.tts_to_file(translated_text,
-          speaker_wav='output_audio.wav',
-          file_path="output_synth.wav",
-          language= self.tran_code
-      )
-      return "output_synth.wav"
-
-  def translate_video(self):
-    audio_path = self.extract_audio()
-    self.org_language_parameters(self.original_language)
-    self.target_language_parameters(self.target_language)
-    transcript_text = self.transcribe_audio(audio_path)
-    translated_text = self.translate_text(transcript_text)
-    translated_audio_path = self.generate_audio(translated_text)
-      #Generate video
-    %cd /content/Wav2Lip
-
-      #This is the detection box padding, if you see it doesnt sit quite right, just adjust the values a bit. Usually the bottom one is the biggest issue
-    pad_top =  0
-    pad_bottom =  15
-    pad_left =  0
-    pad_right =  0
-    rescaleFactor =  1
-    video_path_fix = f"'../{video_path}'"
-    audio_path_fix = f"'../{translated_audio_path}'"
-    !python inference.py --checkpoint_path 'checkpoints/wav2lip_gan.pth' --face $video_path_fix --audio $audio_path_fix --pads $pad_top $pad_bottom $pad_left $pad_right --resize_factor $rescaleFactor --nosmooth --outfile '/content/output_video.mp4'
-
-# Example usage:
-%cd /content/
-video_path = "video.mp4"
-target_language = "German"
-original_language = "English"
-translator = translation(video_path,original_language,target_language)
-translator.translate_video()
-if __name__ == "__main__":
- pass
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+    def extract_audio(self):
+        video = mp.VideoFileClip(self.video_path)
+        audio = video.audio
+        audio_path = "output_audio.wav"
+        audio.write_audiofile(audio_path)
+        st.success("Audio extracted successfully!")
+        return audio_path
+
+    def transcribe_audio(self, audio_path):
+        aai.settings.api_key = "c29eb650444a4ae4bea01d82dd861cbb"
+        config = aai.TranscriptionConfig(language_code=self.lan_code)
+        transcriber = aai.Transcriber(config=config)
+        transcript = transcriber.transcribe(audio_path)
+        transcript_text = transcript.text
+        return transcript_text
+
+    def translate_text(self, transcript_text):
+        base_url = "https://api.cognitive.microsofttranslator.com"
+        endpoint = "/translate"
+        headers = {
+            "Ocp-Apim-Subscription-Key": "cd226bb1f3644276bea01d82dd861cbb",
+            "Content-Type": "application/json",
+            "Ocp-Apim-Subscription-Region": "southeastasia"
+        }
+        params = {
+            "api-version": "3.0",
+            "from": self.lan_code,
+            "to": self.tran_code
+        }
+        body = [{"text": transcript_text}]
+        response = requests.post(base_url + endpoint, headers=headers, params=params, json=body)
+        response.raise_for_status()
+        translation = response.json()[0]["translations"][0]["text"]
+        return translation
+
+    def generate_audio(self, translated_text):
+        tts.tts_to_file(translated_text,
+                        speaker_wav='output_audio.wav',
+                        file_path="output_synth.wav",
+                        language=self.tran_code
+                        )
+        return "output_synth.wav"
+
+    def translate_video(self):
+        audio_path = self.extract_audio()
+        self.org_language_parameters(self.original_language)
+        self.target_language_parameters(self.target_language)
+        transcript_text = self.transcribe_audio(audio_path)
+        translated_text = self.translate_text(transcript_text)
+        translated_audio_path = self.generate_audio(translated_text)
+
+        pad_top = 0
+        pad_bottom = 15
+        pad_left = 0
+        pad_right = 0
+        rescaleFactor = 1
+        video_path_fix = f"'{self.video_path}'"
+        audio_path_fix = f"'{translated_audio_path}'"
+        subprocess.run(['python', 'inference.py', '--checkpoint_path', 'checkpoints/wav2lip_gan.pth',
+                        '--face', video_path_fix, '--audio', audio_path_fix, '--pads',
+                        str(pad_top), str(pad_bottom), str(pad_left), str(pad_right),
+                        '--resize_factor', str(rescaleFactor), '--nosmooth', '--outfile', 'output_video.mp4'])
+
+
+st.title("Translate Your Video")
+st.write("Upload your video and select the original and target languages.")
+
+# Upload video
+video_file = st.file_uploader("Upload Video", type=["mp4"])
+
+if video_file is not None:
+    # Get original and target languages
+    original_language = st.selectbox("Select Original Language", ['English', 'German', 'French', 'Spanish'])
+    target_language = st.selectbox("Select Target Language", ['English', 'German', 'French', 'Spanish'])
+
+    translation = Translation(video_path=video_file.name,
+                              original_language=original_language,
+                              target_language=target_language)
+
+    if st.button("Translate"):
+        translation.translate_video()
+        st.success("Video translation complete! You can download the translated video.")