tee342 commited on
Commit
acc2409
·
verified ·
1 Parent(s): b4e6504

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +87 -51
app.py CHANGED
@@ -17,7 +17,6 @@ from PIL import Image
17
  import zipfile
18
  import datetime
19
  import librosa
20
- import joblib
21
  import warnings
22
  from faster_whisper import WhisperModel
23
  from mutagen.mp3 import MP3
@@ -267,7 +266,7 @@ def batch_process_audio(files, selected_effects, isolate_vocals, preset_name, ex
267
  except Exception as e:
268
  return None, f"❌ Batch processing failed: {str(e)}"
269
 
270
- # === Whisper Transcription Tab ===
271
  whisper_model = WhisperModel("base")
272
 
273
  def transcribe_audio(audio_path):
@@ -353,10 +352,42 @@ def mix_tracks(track1, track2, volume_offset=0):
353
 
354
  # === Voice Style Transfer (Dummy) ===
355
  def apply_style_transfer(audio_path, mood="Happy"):
356
- # Replace with real model later
357
  return audio_path
358
 
359
- # === UI Setup ===
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
360
  effect_options = [
361
  "Noise Reduction",
362
  "Compress Dynamic Range",
@@ -436,13 +467,13 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
436
  clear_btn=None
437
  )
438
 
439
- # --- Transcribe & Edit Tab ===
440
  with gr.Tab("📝 Transcribe & Edit"):
441
  gr.Interface(
442
  fn=transcribe_audio,
443
  inputs=gr.Audio(label="Upload Audio", type="filepath"),
444
  outputs=gr.Textbox(label="Transcribed Text", lines=10),
445
- title="Transcribe Spoken Content",
446
  description="Convert voice to text and edit it before exporting again."
447
  )
448
 
@@ -473,58 +504,50 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
473
  )
474
 
475
  # --- Voice Cloning (Dummy) ===
476
- def clone_voice(*args):
477
- return "⚠️ Voice cloning requires additional setup"
478
-
479
  with gr.Tab("🎭 Voice Cloning (Dubbing)"):
480
  gr.Interface(
481
- fn=clone_voice,
482
  inputs=[
483
  gr.File(label="Source Voice Clip"),
484
- gr.File(label="Target Voice Clip"),
485
- gr.Textbox(label="Text to Clone", lines=5)
486
  ],
487
- outputs=gr.Audio(label="Cloned Output", type="filepath"),
488
- title="Replace One Voice With Another",
489
  description="Clone voice from source to target speaker using AI"
490
  )
491
 
492
  # --- Auto-Save / Resume Sessions ===
493
- session_state = gr.State()
494
-
495
- def auto_save_session(audio, preset, effects):
496
- return {"audio": audio, "preset": preset, "effects": effects}
497
-
498
- def auto_load_session(session):
499
- if session and "audio" in session:
500
- return session["audio"], session["preset"], session["effects"]
501
- return None, None, None
502
-
503
  with gr.Tab("🧾 Auto-Save & Resume"):
504
- gr.Interface(
505
- fn=auto_save_session,
506
- inputs=[
507
- gr.Audio(label="Upload Audio", type="filepath"),
508
- gr.Dropdown(choices=preset_names, label="Used Preset", value=preset_names[0]),
509
- gr.CheckboxGroup(choices=effect_options, label="Applied Effects")
510
- ],
511
- outputs=session_state,
512
- title="Auto-Save Your Session",
513
- description="Save your current state and resume editing later",
514
- allow_flagging="never"
515
- )
516
-
517
- gr.Interface(
518
- fn=auto_load_session,
519
- inputs=session_state,
520
- outputs=[
521
- gr.Audio(label="Loaded Audio", type="filepath"),
522
- gr.Dropdown(choices=preset_names, label="Loaded Preset"),
523
- gr.CheckboxGroup(choices=effect_options, label="Loaded Effects")
524
- ],
525
- title="Resume Last Session",
526
- description="Reload your last edit state",
527
- allow_flagging="never"
 
 
 
 
 
 
528
  )
529
 
530
  # --- VAD – Detect & Remove Silence ===
@@ -547,7 +570,7 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
547
  fn=save_project,
548
  inputs=[
549
  gr.File(label="Original Audio"),
550
- gr.Dropdown(choices=preset_names, label="Used Preset", value=preset_names[0]),
551
  gr.CheckboxGroup(choices=effect_options, label="Applied Effects")
552
  ],
553
  outputs=gr.File(label="Project File (.aiproj)"),
@@ -577,10 +600,10 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
577
  ],
578
  outputs=gr.File(label="Mixed Output"),
579
  title="Overlay Two Tracks",
580
- description="Mix or subtract two audio files."
581
  )
582
 
583
- # --- Voice Style Transfer (Dummy) ===
584
  with gr.Tab("🧠 Voice Style Transfer"):
585
  gr.Interface(
586
  fn=apply_style_transfer,
@@ -593,4 +616,17 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
593
  description="Shift the emotional style of any voice clip."
594
  )
595
 
 
 
 
 
 
 
 
 
 
 
 
 
 
596
  demo.launch()
 
17
  import zipfile
18
  import datetime
19
  import librosa
 
20
  import warnings
21
  from faster_whisper import WhisperModel
22
  from mutagen.mp3 import MP3
 
266
  except Exception as e:
267
  return None, f"❌ Batch processing failed: {str(e)}"
268
 
269
+ # === Transcribe & Edit Tab ===
270
  whisper_model = WhisperModel("base")
271
 
272
  def transcribe_audio(audio_path):
 
352
 
353
  # === Voice Style Transfer (Dummy) ===
354
  def apply_style_transfer(audio_path, mood="Happy"):
 
355
  return audio_path
356
 
357
+ # === Load Custom Plugins ===
358
+ def load_plugin(plugin_file, audio_file):
359
+ try:
360
+ from importlib.util import spec_from_file_location, module_from_spec
361
+ spec = spec_from_file_location("plugin", plugin_file.name)
362
+ plugin = module_from_spec(spec)
363
+ spec.loader.exec_module(plugin)
364
+
365
+ # Run plugin
366
+ audio = AudioSegment.from_file(audio_file)
367
+ processed = plugin.process(audio)
368
+
369
+ out_path = os.path.join(tempfile.gettempdir(), "plugin_output.wav")
370
+ processed.export(out_path, format="wav")
371
+ return out_path
372
+ except Exception as e:
373
+ return f"⚠️ Plugin error: {str(e)}"
374
+
375
+ # === Auto-Save / Resume Session Handling ===
376
+ session_state = gr.State()
377
+
378
+ def save_or_resume_session(audio, preset, effects, action="save"):
379
+ if action == "save":
380
+ return {"audio": audio, "preset": preset, "effects": effects}, None, None, None
381
+ elif action == "load" and audio is not None:
382
+ return (
383
+ None,
384
+ audio.get("audio"),
385
+ audio.get("preset"),
386
+ audio.get("effects")
387
+ )
388
+ return None, None, None, None
389
+
390
+ # === UI ===
391
  effect_options = [
392
  "Noise Reduction",
393
  "Compress Dynamic Range",
 
467
  clear_btn=None
468
  )
469
 
470
+ # --- Transcribe & Edit ---
471
  with gr.Tab("📝 Transcribe & Edit"):
472
  gr.Interface(
473
  fn=transcribe_audio,
474
  inputs=gr.Audio(label="Upload Audio", type="filepath"),
475
  outputs=gr.Textbox(label="Transcribed Text", lines=10),
476
+ title="Transcribe & Edit Spoken Content",
477
  description="Convert voice to text and edit it before exporting again."
478
  )
479
 
 
504
  )
505
 
506
  # --- Voice Cloning (Dummy) ===
 
 
 
507
  with gr.Tab("🎭 Voice Cloning (Dubbing)"):
508
  gr.Interface(
509
+ fn=apply_style_transfer,
510
  inputs=[
511
  gr.File(label="Source Voice Clip"),
512
+ gr.Radio(["Happy", "Sad", "Angry", "Calm"], label="Choose Tone")
 
513
  ],
514
+ outputs=gr.Audio(label="Stylized Output", type="filepath"),
515
+ title="Replace Voice Using AI",
516
  description="Clone voice from source to target speaker using AI"
517
  )
518
 
519
  # --- Auto-Save / Resume Sessions ===
 
 
 
 
 
 
 
 
 
 
520
  with gr.Tab("🧾 Auto-Save & Resume"):
521
+ gr.Markdown("Save your current session and come back later.")
522
+
523
+ action_radio = gr.Radio(["save", "load"], label="Action", value="save")
524
+ audio_input = gr.Audio(label="Upload or Load Audio", type="filepath")
525
+ preset_dropdown = gr.Dropdown(choices=preset_names, label="Used Preset", value=preset_names[0] if preset_names else None)
526
+ effect_checkbox = gr.CheckboxGroup(choices=effect_options, label="Applied Effects")
527
+ save_btn = gr.Button("Save or Load Session")
528
+
529
+ session_data = gr.State()
530
+
531
+ loaded_audio = gr.Audio(label="Loaded Audio", type="filepath")
532
+ loaded_preset = gr.Dropdown(choices=preset_names, label="Loaded Preset")
533
+ loaded_effects = gr.CheckboxGroup(choices=effect_options, label="Loaded Effects")
534
+
535
+ def handle_action(audio, preset, effects, action):
536
+ if action == "save":
537
+ return {"audio": audio, "preset": preset, "effects": effects}, None, None, None
538
+ elif action == "load" and isinstance(audio, dict):
539
+ return (
540
+ None,
541
+ audio.get("audio"),
542
+ audio.get("preset"),
543
+ audio.get("effects")
544
+ )
545
+ return None, None, None, None
546
+
547
+ save_btn.click(
548
+ fn=handle_action,
549
+ inputs=[audio_input, preset_dropdown, effect_checkbox, action_radio],
550
+ outputs=[session_data, loaded_audio, loaded_preset, loaded_effects]
551
  )
552
 
553
  # --- VAD – Detect & Remove Silence ===
 
570
  fn=save_project,
571
  inputs=[
572
  gr.File(label="Original Audio"),
573
+ gr.Dropdown(choices=preset_names, label="Used Preset", value=preset_names[0] if preset_names else None),
574
  gr.CheckboxGroup(choices=effect_options, label="Applied Effects")
575
  ],
576
  outputs=gr.File(label="Project File (.aiproj)"),
 
600
  ],
601
  outputs=gr.File(label="Mixed Output"),
602
  title="Overlay Two Tracks",
603
+ description="Mix, blend, or subtract two audio files."
604
  )
605
 
606
+ # --- Voice Style Transfer ===
607
  with gr.Tab("🧠 Voice Style Transfer"):
608
  gr.Interface(
609
  fn=apply_style_transfer,
 
616
  description="Shift the emotional style of any voice clip."
617
  )
618
 
619
+ # --- Custom Effect Plugin System ===
620
+ with gr.Tab("🧩 Load Custom Effect"):
621
+ gr.Interface(
622
+ fn=load_plugin,
623
+ inputs=[
624
+ gr.File(label="Upload .py plugin"),
625
+ gr.Audio(label="Upload Audio", type="filepath")
626
+ ],
627
+ outputs=gr.Audio(label="Processed Output", type="filepath"),
628
+ title="Run Your Own Python Effect",
629
+ description="Upload a .py file with a 'process' function"
630
+ )
631
+
632
  demo.launch()