Update app.py

app.py

@@ -4,6 +4,7 @@ from deep_translator import GoogleTranslator
 from transformers import pipeline
 import os
 from gtts import gTTS
+from pytube import YouTube
 
 pipe = pipeline(model="freeja/lab2-whisper-sv")
 
@@ -32,13 +33,13 @@ def text_to_speech(text,language):
     gTTS(text,lang,slow=False)
 
 def transcribe_video(URL):
-    video = Youtube(URL)
+    video = YouTube(URL)
     yt = video.streams.get_audio_only()
     yt.download()
     text = pipe(yt)["text"]
     return text
 
-iface = gr.Interface(
+"""iface = gr.Interface(
     fn=transcribe_audio,
     inputs=[
         gr.Audio(source="microphone", type="filepath", label="Transcribe from Microphone"),
@@ -46,6 +47,27 @@ iface = gr.Interface(
     ],
     outputs="text",
     title="Whisper Small Swedish",
-    description="Realtime demo for Swedish speech recognition with translation using a fine-tuned Whisper small model")
+    description="Realtime demo for Swedish speech recognition with translation using a fine-tuned Whisper small model")"""
+
+video_transcription = gr.Interface(
+    fn = transcribe_video, 
+    inputs = "text", 
+    outputs = "text",
+    title = "Whisper Small Swedish",
+    description = "Realtime demo for Swedish speech recognition with translation using a fine-tuned Whisper small model"",
+)
+
+audio_transcription = gr.Interface(
+    fn=transcribe_audio, 
+    inputs=[
+        gr.Audio(source="microphone", type="filepath", label="Transcribe from Microphone"),
+        gr.Dropdown(["English","Spanish","Dutch","French","Italian"], value="English", label="Translate to ")
+    ], 
+    outputs="text",
+    title="Whisper Small Swedish",
+    description="Realtime demo for Swedish speech recognition with translation using a fine-tuned Whisper small model"",
+)
+
+iface = gr.TabbedInterface([audio_transcription, video_transcription], ["Transcribe Audio", "Transcribe Video"])
 
 iface.launch()