rewored webapp

autotranscript/app/gradio_app.py

@@ -15,8 +15,34 @@ Usage:
     
 """
 
-from autotranscript import AutoTranscribe
+"""
+Gradio Audio Transcription App.
+--------------------------------
+
+This module provides an interface to transcribe audio files using the 
+AutoTranscribe model. Users can either upload an audio file or record their speech 
+live for transcription. The application supports multiple languages and provides 
+options to specify the number of speakers and the language of the audio.
+
+Attributes:
+    LANGUAGES (list): A list of supported languages for transcription.
+
+Usage:
+    Run this script to start the Gradio web interface for audio transcription.
+    
+"""
+
+import json
+
 import gradio as gr
+from autotranscript import AutoTranscribe, Transcript
+
+
+theme = gr.themes.Soft(
+    primary_hue="green",
+    secondary_hue='orange',
+    neutral_hue="gray",  
+)
 
 LANGUAGES = [
     "Afrikaans", "Arabic", "Armenian", "Azerbaijani", "Belarusian",
@@ -33,70 +59,282 @@ LANGUAGES = [
     "Vietnamese", "Welsh"
 ]
 
-
+class GradioTranscriptionInterface:
-def gradio_server(model : AutoTranscribe):
     """
-    Sets up and launches the Gradio interface for audio transcription.
+    Interface handling the interaction between Gradio UI and the Audio Transcription system.
-
-    Args:
-        model (AutoTranscribe): An instance of the AutoTranscribe model for transcription.
     """
-    def transcribe(audio, microphone, number_of_speakers, language):
+
+    def __init__(self, model: AutoTranscribe):
         """
-        Transcribes the provided audio input based on the given parameters.
+        Initializes the GradioTranscriptionInterface with a transcription model.
 
         Args:
-            audio (str): Filepath to the uploaded audio file.
+            model (AutoTranscribe): Model responsible for audio transcription tasks.
-            microphone (str): Filepath to the recorded audio.
+        """
-            number_of_speakers (int): Number of speakers in the audio.
+        self.model = model
-            language (str): Language of the audio content.
+
+    def auto_transcribe(self, source,
+                        num_speakers : int,
+                        translation : bool,
+                        language : str):
+        """
+        Shortcut method for the AutoTranscribe task.
 
         Returns:
             tuple: Transcribed text (str), JSON output (dict)
         """
-        kwargs = {}
-        if number_of_speakers != 0:
-            kwargs["num_speakers"] = number_of_speakers
-        if language != "None":
-            kwargs["language"] = language
         
-        print()
+        kwargs = {
+            "num_speakers": num_speakers if num_speakers != 0 else None,
+            "language": language if language != "None" else None,
+            "task": 'translate' if translation else None
+        }
         
-        if audio is not None:
+        try:
-            out = model.transcribe(audio, **kwargs)
+            result = self.model.autotranscribe(source, **kwargs)
-        elif microphone is not None:
+        except ValueError:
-            out = model.transcribe(microphone , **kwargs)
+            raise gr.Error("Couldn't detect any speech in the provided audio. \
-        else:
+                    Please try again!")
-            out = "Please upload an audio file or record one."
+        return str(result), result.get_json()
 
-        return str(out), out.get_json(), out.get_md()
 
-    gr.Interface(
+    def transcribe(self, source, translation, language):
-        fn=transcribe, 
+        """
-        inputs=[
+        Shortcut method for the Transcribe task.
-            gr.Audio(source= "upload", type="filepath", label="Upload Your Audio File",
+
-                     interactive=True),
+        Returns:
-            gr.Audio(source= "microphone", type="filepath", label="Record Your Audio",
+            str: Transcribed text.
-                     interactive=True, container= False),
+        """
-            gr.Number(value=0, label= "Number of speakers (optional)", 
+        kwargs = {
-                      info = "Number of speakers in the audio file. If you don't know, leave it at 0."), 
+            "language": language if language != "None" else None,
-            gr.Dropdown(LANGUAGES,
+            "task": 'translate' if translation == "Yes" else None
-                        label="Language (optional)", value = "None",
+        }
-                        info="Language of the audio file. If you don't know, leave it at None.")
+        
-        ],
+        result = self.model.transcribe(source, **kwargs)
-        outputs=[
+        return str(result)
-            gr.Textbox(label="Transcription"),
+
-            gr.JSON(label="Raw Output", container= False),
+    def perform_diarisation(self, source, num_speakers):
-        ],
+        """
-        title="Audio Transcription",
+        Shortcut method for the Diarisation task.
-        description="Upload an audio file to transcribe its content. Powered by AutoTranscribe!",
+
-        theme="soft",       # Example of a more modern theme
+        Returns:
-        server_port=7860,
+            str: JSON output of diarisation result.
-        server_name="autotranscribe",   
+        """
-    ).queue().launch() 
+        kwargs = {
+            "num_speakers": num_speakers if num_speakers != 0 else None,
+        }
+        
+        
+        try:
+            result = self.model.diarization(source, **kwargs)
+        except ValueError:
+            raise gr.Error("Couldn't detect any speech in the provided audio. \
+                    Please try again!")
+        return json.dumps(result, indent=2)
+
+####
+# Gradio Interface
+####
+
+def gradio_Interface(model : AutoTranscribe = None):
+    
+    if model is None:
+        model = AutoTranscribe()
+        
+    pipe = GradioTranscriptionInterface(model)
+
+    def select_task(choice):
+        if choice == 'Auto Transcribe':
+            
+            return (gr.update(visible = True),
+                    gr.update(visible = True),
+                    gr.update(visible = True))
+                    
+            
+        elif choice == 'Transcribe':
+            
+            return (gr.update(visible = False),
+                    gr.update(visible = True),
+                    gr.update(visible = True))
+
+            
+        elif choice == 'Diarisation':
+            
+            return (gr.update(visible = True),
+                    gr.update(visible = False),
+                    gr.update(visible = False))
+        
+    def select_origin(choice):
+        if choice == "Upload Audio":
+            
+            return (gr.update(visible = True),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None))
+        
+        elif choice == "Record Audio":
+            
+            return (gr.update(visible = False, value = None),
+                    gr.update(visible = True),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None))
+
+        elif choice == "Upload Video":
+            
+            return (gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = True),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None))
+        
+        elif choice == "Record Video":
+            
+            return (gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = True),
+                    gr.update(visible = False, value = None))
+            
+        elif choice == "File":
+            
+            return (gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = False, value = None),
+                    gr.update(visible = True))
+
+    def run_scribe(task, num_speakers, translate, language, audio1, audio2, video1, video2, file_in, progress = gr.Progress(track_tqdm= True)):
+        # get *args which are not None
+        progress(0, desc='Starting task...')
+        source = audio1 or audio2 or video1 or video2 or file_in
+        
+        if task == 'Auto Transcribe':
+            
+            out_str , out_json = pipe.auto_transcribe(source = source,
+                                num_speakers = num_speakers,
+                                translation = translate,
+                                language = language)
+            
+            return (gr.update(value = out_str, visible = True),
+                    gr.update(value = out_json, visible = True),
+                    gr.update(visible = True),
+                    gr.update(visible = True))        
+            
+        elif task == 'Transcribe':
+            
+            out = pipe.transcribe(source = source,
+                                translation = translate,
+                                language = language)
+            
+            return (gr.update(value = out, visible = True),
+                    gr.update(value = None, visible = False),
+                    gr.update(visible = False),
+                    gr.update(visible = False))
+            
+        elif task == 'Diarisation':
+            
+            out = pipe.perform_diarisation(source = source,
+                                num_speakers = num_speakers)
+            
+            return (gr.update(value = None, visible = False),
+                    gr.update(value = out, visible = True),
+                    gr.update(visible = False),
+                    gr.update(visible = False))
+        
+    def annotate_output(annoation : str, out_json : dict):
+        # get *args which are not None
+        
+        trans = Transcript.from_json(out_json)
+        trans = trans.annotate(*annoation.split(","))
+
+        return gr.update(value = str(trans)),gr.update(value = trans.get_json())
+        
+        
+    with gr.Blocks(theme=theme,title='ScrAIbe: Automatic Audio Transcription') as demo:
+            
+        # Define components
+        header = open("header.html", "r").read()
+        gr.HTML(header, visible= True, show_label=False)
+        
+        with gr.Row():
+            
+            with gr.Column():
+            
+                task = gr.Radio(["Auto Transcribe", "Transcribe", "Diarisation"], label="Task",
+                                value= 'Auto Transcribe')
+                
+                num_speakers = gr.Number(value=0, label= "Number of speakers (optional)", 
+                                info = "Number of speakers in the audio file. If you don't know,\
+                                    leave it at 0.", visible= True)
+                
+                translate = gr.Checkbox(label="Translation", choices=[True, False], value = False,
+                                info="Select 'Yes' to have the output translated into English.",
+                                visible= True)
+                
+                language = gr.Dropdown(LANGUAGES,
+                                label="Language (optional)", value = "None",
+                                info="Language of the audio file. If you don't know,\
+                                    leave it at None.", visible= True)
+                
+                input = gr.Radio(["Upload Audio", "Record Audio", "Upload Video","Record Video" 
+                                    ,"File"], label="Input Type", value="Upload Audio")
+                
+                audio1 = gr.Audio(source="upload", type="filepath", label="Upload Audio",
+                                    interactive= True, visible= True)
+                audio2 = gr.Audio(source="microphone", label="Record Audio", type="filepath",
+                                    interactive= True, visible= False)
+                video1 = gr.Video(source="upload", type="filepath", label="Upload Video",
+                                    interactive= True, visible= False)
+                video2 = gr.Video(source="webcam", label="Record Video", type="filepath",
+                                    interactive= True, visible= False)
+                file_in = gr.File(label="Upload File", interactive= True, visible= False)
+                
+                submit = gr.Button()
+            
+            with gr.Column():
+                
+                out_txt = gr.Textbox(label="Output",
+                                        visible= True, show_copy_button=True)
+                
+                out_json = gr.JSON(label="JSON Output",
+                                    visible= False, show_copy_button=True)
+                
+                annoation = gr.Textbox(label="Name your speaker's",
+                                    info= "Please provide a list of the speakers arranged \
+                                    in the order in which they appear in the input. Use comma ',' \
+                                    as a seperator. Be aware that the first name is given \
+                                        to SPEAKER_00 the second to SPEAKER_01 and so on.",
+                                    visible= False, interactive= True)
+                
+                annotate = gr.Button(value="Annotate", visible= False, interactive= True)
+            
+        # Define usage of components
+        input.change(fn=select_origin, inputs=[input],
+                        outputs=[audio1, audio2, video1, video2, file_in])
+        
+        task.change(fn=select_task, inputs=[task],
+                    outputs=[num_speakers, translate, language])
+        
+        translate.change(fn= lambda x : gr.update(value = x),
+                            inputs=[translate], outputs=[translate])
+        num_speakers.change(fn= lambda x : gr.update(value = x),
+                            inputs=[num_speakers], outputs=[num_speakers])
+        language.change(fn= lambda x : gr.update(value = x), 
+                        inputs=[language], outputs=[language])
+        
+        submit.click(fn = run_scribe, 
+                        inputs=[task, num_speakers, translate, language, audio1,
+                                audio2, video1, video2, file_in],
+                        outputs=[out_txt, out_json, annoation, annotate])
+        
+        annotate.click(fn = annotate_output, inputs=[annoation, out_json],
+                        outputs=[out_txt, out_json])
+        
+    return demo
 
     
 if __name__ == "__main__":
     
-    model = AutoTranscribe()
+    gradio_Interface().queue().launch()
-    gradio_server(model)

autotranscript/app/header.html

@@ -0,0 +1,66 @@
+<!-- Importing Cormorant Garamond font from Google Fonts -->
+<link href="https://fonts.googleapis.com/css2?family=Cormorant+Garamond:wght@400;700&display=swap" rel="stylesheet">
+
+<style>
+    .header-container {
+        display: flex;
+        align-items: center;
+        justify-content: center;
+        position: relative;
+        padding-top: 30px;
+    }
+    .logo-container {
+        position: absolute;
+        top: 50%;
+        right: 20px;
+        transform: translateY(-50%);
+        width: 300px;
+    }
+    .logo {
+        width: 100%;
+        height: auto;
+    }
+    h1 {
+        font-family: 'Cormorant Garamond', serif;
+        font-size: 50px !important; /* Increased font size */
+        font-weight: bold;
+        color: #50AF31;
+        margin: 0;
+        position: relative;
+        padding: 0.5em 0;
+    }
+    h1::before, h1::after {
+        content: "";
+        position: absolute;
+        height: 2px;
+        width: 80%;
+        background-color: #50AF31;
+        left: 10%;
+    }
+    h1::before {
+        top: 0.5em;
+    }
+    h1::after {
+        bottom: 0.5em;
+    }
+    p, h2 {
+        font-size: 16px;
+        margin: 10px 0;
+        line-height: 1.4;
+    }
+</style>
+
+<div class="header-container">
+    <h1>ScrAIbe</h1>
+    <div class="logo-container">
+        <a href="https://www.kida-bmel.de/"> <!-- Replace with your actual URL -->
+            <img src="file/Logo_KIDA_bmel_green.svg" alt="KIDA Logo" class="logo">
+        </a>
+    </div>
+</div>
+<div style="text-align: center; padding: 20px 10%;">
+    <p>
+        Upload, record, or provide a video with audio for transcription. Our toolkit is designed to transcribe content from multiple languages accurately. The integrated speaker diarisation feature identifies different speakers, ensuring a smooth transcription experience. For optimal results, indicate the number of speakers and the original language of the content.
+    </p>
+    <h2 style="font-weight: bold; color: #50AF31;">What would you like to do next?</h2>
+</div>

autotranscript/cli.py

@@ -11,6 +11,7 @@ import json
 from sympy import use
 
 from .autotranscript import AutoTranscribe
+from .app.gradio_app import gradio_Interface
 
 from whisper.tokenizer import LANGUAGES , TO_LANGUAGE_CODE
 from torch.cuda import is_available
@@ -160,7 +161,7 @@ def cli():
                 
     
     if start_server: # unfinished code
-        from .app.gradio_app import gradio_Interface
+        
         gradio_Interface(model).queue().launch(server_port=args.port, server_name=args.server_name)
         
 

autotranscript/transcript_exporter.py

@@ -1,5 +1,6 @@
 import json
 import time
+from traceback import print_stack
 
 
 from typing import Union
@@ -50,6 +51,7 @@ class Transcript:
         
         if args:
             for arg, speaker in zip(args, sorted(self.speakers)):
+                
                 annotations[speaker] = arg
         
         invalid_speakers = set(kwargs.keys()) - set(self.speakers)

setup.py

@@ -40,6 +40,7 @@ if __name__ == "__main__":
         author='Jacob Schmieder',
         author_email='Jacob.Schmieder@dbfz.de',
         description='Transcription tool for audio files based on Whisper and Pyannote',
+        package_data={ "header" : ["app/header.html"], "logo" : ["app/Logo_KIDA_bmel_green.svg"]},
         entry_points={'console_scripts':
-            ['autotranscript = autotranscript.autotranscript:cli']}
+            ['autotranscript = autotranscript.cli:cli']}
     )