Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -23,6 +23,7 @@ from faster_whisper import WhisperModel
|
|
| 23 |
from mutagen.mp3 import MP3
|
| 24 |
from mutagen.id3 import ID3, TIT2, TPE1, TALB, TYER
|
| 25 |
from TTS.api import TTS
|
|
|
|
| 26 |
|
| 27 |
# Suppress warnings
|
| 28 |
warnings.filterwarnings("ignore")
|
|
@@ -318,6 +319,11 @@ def save_project(audio_path, preset_name, effects):
|
|
| 318 |
pickle.dump(project_data, f)
|
| 319 |
return out_path
|
| 320 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 321 |
# === Trim Silence Automatically (VAD) ===
|
| 322 |
def detect_silence(audio_file, silence_threshold=-50.0, min_silence_len=1000):
|
| 323 |
audio = AudioSegment.from_file(audio_file)
|
|
@@ -345,25 +351,12 @@ def mix_tracks(track1, track2, volume_offset=0):
|
|
| 345 |
mixed.export(out_path, format="wav")
|
| 346 |
return out_path
|
| 347 |
|
| 348 |
-
# ===
|
| 349 |
-
def
|
| 350 |
-
|
| 351 |
-
|
| 352 |
-
spec = spec_from_file_location("plugin", plugin_file.name)
|
| 353 |
-
plugin = module_from_spec(spec)
|
| 354 |
-
spec.loader.exec_module(plugin)
|
| 355 |
-
|
| 356 |
-
# Run plugin
|
| 357 |
-
audio = AudioSegment.from_file(audio_file)
|
| 358 |
-
processed = plugin.process(audio)
|
| 359 |
-
|
| 360 |
-
out_path = os.path.join(tempfile.gettempdir(), "plugin_output.wav")
|
| 361 |
-
processed.export(out_path, format="wav")
|
| 362 |
-
return out_path
|
| 363 |
-
except Exception as e:
|
| 364 |
-
return f"β οΈ Plugin error: {str(e)}"
|
| 365 |
|
| 366 |
-
# === UI ===
|
| 367 |
effect_options = [
|
| 368 |
"Noise Reduction",
|
| 369 |
"Compress Dynamic Range",
|
|
@@ -443,13 +436,13 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 443 |
clear_btn=None
|
| 444 |
)
|
| 445 |
|
| 446 |
-
# --- Transcribe & Edit
|
| 447 |
with gr.Tab("π Transcribe & Edit"):
|
| 448 |
gr.Interface(
|
| 449 |
fn=transcribe_audio,
|
| 450 |
inputs=gr.Audio(label="Upload Audio", type="filepath"),
|
| 451 |
outputs=gr.Textbox(label="Transcribed Text", lines=10),
|
| 452 |
-
title="Transcribe
|
| 453 |
description="Convert voice to text and edit it before exporting again."
|
| 454 |
)
|
| 455 |
|
|
@@ -479,28 +472,62 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 479 |
description="Detect speaker similarity using AI."
|
| 480 |
)
|
| 481 |
|
| 482 |
-
# ---
|
| 483 |
-
def
|
| 484 |
-
|
| 485 |
-
import json
|
| 486 |
-
data = {"effects": selected_effects, "preset": preset_name, "format": export_format}
|
| 487 |
-
encoded = base64.b64encode(json.dumps(data).encode()).decode()
|
| 488 |
-
return f"https://your-space-url?preset={encoded}"
|
| 489 |
|
| 490 |
-
with gr.Tab("
|
| 491 |
gr.Interface(
|
| 492 |
-
fn=
|
| 493 |
inputs=[
|
| 494 |
-
gr.
|
| 495 |
-
gr.
|
| 496 |
-
gr.
|
| 497 |
],
|
| 498 |
-
outputs=gr.
|
| 499 |
-
title="
|
| 500 |
-
description="
|
| 501 |
)
|
| 502 |
|
| 503 |
-
# ---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 504 |
with gr.Tab("βοΈ Trim Silence Automatically"):
|
| 505 |
gr.Interface(
|
| 506 |
fn=detect_silence,
|
|
@@ -511,10 +538,10 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 511 |
],
|
| 512 |
outputs=gr.File(label="Trimmed Output"),
|
| 513 |
title="Auto-Detect & Remove Silence",
|
| 514 |
-
description="
|
| 515 |
)
|
| 516 |
|
| 517 |
-
# --- Load/Save Project ===
|
| 518 |
with gr.Tab("π Save/Load Project"):
|
| 519 |
gr.Interface(
|
| 520 |
fn=save_project,
|
|
@@ -528,6 +555,17 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 528 |
description="Save your session, effects, and settings in one file to reuse later."
|
| 529 |
)
|
| 530 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 531 |
# --- Mix Two Tracks ===
|
| 532 |
with gr.Tab("π Mix Two Tracks"):
|
| 533 |
gr.Interface(
|
|
@@ -542,17 +580,17 @@ with gr.Blocks(title="AI Audio Studio", css="style.css") as demo:
|
|
| 542 |
description="Mix or subtract two audio files."
|
| 543 |
)
|
| 544 |
|
| 545 |
-
# ---
|
| 546 |
-
with gr.Tab("
|
| 547 |
gr.Interface(
|
| 548 |
-
fn=
|
| 549 |
inputs=[
|
| 550 |
-
gr.
|
| 551 |
-
gr.
|
| 552 |
],
|
| 553 |
-
outputs=gr.Audio(label="
|
| 554 |
-
title="
|
| 555 |
-
description="
|
| 556 |
)
|
| 557 |
|
| 558 |
demo.launch()
|
|
|
|
| 23 |
from mutagen.mp3 import MP3
|
| 24 |
from mutagen.id3 import ID3, TIT2, TPE1, TALB, TYER
|
| 25 |
from TTS.api import TTS
|
| 26 |
+
import pickle
|
| 27 |
|
| 28 |
# Suppress warnings
|
| 29 |
warnings.filterwarnings("ignore")
|
|
|
|
| 319 |
pickle.dump(project_data, f)
|
| 320 |
return out_path
|
| 321 |
|
| 322 |
+
def load_project(project_file):
|
| 323 |
+
with open(project_file.name, "rb") as f:
|
| 324 |
+
data = pickle.load(f)
|
| 325 |
+
return data["preset"], data["effects"]
|
| 326 |
+
|
| 327 |
# === Trim Silence Automatically (VAD) ===
|
| 328 |
def detect_silence(audio_file, silence_threshold=-50.0, min_silence_len=1000):
|
| 329 |
audio = AudioSegment.from_file(audio_file)
|
|
|
|
| 351 |
mixed.export(out_path, format="wav")
|
| 352 |
return out_path
|
| 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 |
clear_btn=None
|
| 437 |
)
|
| 438 |
|
| 439 |
+
# --- Transcribe & Edit Tab ===
|
| 440 |
with gr.Tab("π 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 |
|
|
|
|
| 472 |
description="Detect speaker similarity using AI."
|
| 473 |
)
|
| 474 |
|
| 475 |
+
# --- Voice Cloning (Dummy) ===
|
| 476 |
+
def clone_voice(*args):
|
| 477 |
+
return "β οΈ Voice cloning requires additional setup"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 478 |
|
| 479 |
+
with gr.Tab("π 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("π§Ύ 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 β Detect & Remove Silence ===
|
| 531 |
with gr.Tab("βοΈ Trim Silence Automatically"):
|
| 532 |
gr.Interface(
|
| 533 |
fn=detect_silence,
|
|
|
|
| 538 |
],
|
| 539 |
outputs=gr.File(label="Trimmed Output"),
|
| 540 |
title="Auto-Detect & Remove Silence",
|
| 541 |
+
description="Detect and trim silence at start/end or between words"
|
| 542 |
)
|
| 543 |
|
| 544 |
+
# --- Load/Save Project File (.aiproj) ===
|
| 545 |
with gr.Tab("π Save/Load Project"):
|
| 546 |
gr.Interface(
|
| 547 |
fn=save_project,
|
|
|
|
| 555 |
description="Save your session, effects, and settings in one file to reuse later."
|
| 556 |
)
|
| 557 |
|
| 558 |
+
gr.Interface(
|
| 559 |
+
fn=load_project,
|
| 560 |
+
inputs=gr.File(label="Upload .aiproj File"),
|
| 561 |
+
outputs=[
|
| 562 |
+
gr.Dropdown(choices=preset_names, label="Loaded Preset"),
|
| 563 |
+
gr.CheckboxGroup(choices=effect_options, label="Loaded Effects")
|
| 564 |
+
],
|
| 565 |
+
title="Resume Last Project",
|
| 566 |
+
description="Load your saved session"
|
| 567 |
+
)
|
| 568 |
+
|
| 569 |
# --- Mix Two Tracks ===
|
| 570 |
with gr.Tab("π Mix Two Tracks"):
|
| 571 |
gr.Interface(
|
|
|
|
| 580 |
description="Mix or subtract two audio files."
|
| 581 |
)
|
| 582 |
|
| 583 |
+
# --- Voice Style Transfer (Dummy) ===
|
| 584 |
+
with gr.Tab("π§ Voice Style Transfer"):
|
| 585 |
gr.Interface(
|
| 586 |
+
fn=apply_style_transfer,
|
| 587 |
inputs=[
|
| 588 |
+
gr.Audio(label="Upload Voice Clip", type="filepath"),
|
| 589 |
+
gr.Radio(["Happy", "Sad", "Angry", "Calm"], label="Choose Tone")
|
| 590 |
],
|
| 591 |
+
outputs=gr.Audio(label="Stylized Output", type="filepath"),
|
| 592 |
+
title="Change Emotional Tone of Voice",
|
| 593 |
+
description="Shift the emotional style of any voice clip."
|
| 594 |
)
|
| 595 |
|
| 596 |
demo.launch()
|