Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -595,7 +595,119 @@ def parse_point_coordinates_text(coordinates_text: str) -> List[List[int]]:
|
|
| 595 |
def build_interface():
|
| 596 |
with gr.Blocks(title="Multimodal AI Demo", theme=gr.themes.Soft()) as demo_block:
|
| 597 |
gr.Markdown("# AI модели")
|
| 598 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 599 |
with gr.Tab("Детекция объектов"):
|
| 600 |
gr.Markdown("## Детекция объектов")
|
| 601 |
with gr.Row():
|
|
@@ -828,5 +940,4 @@ def build_interface():
|
|
| 828 |
|
| 829 |
if __name__ == "__main__":
|
| 830 |
interface_block = build_interface()
|
| 831 |
-
interface_block.launch(share=True)
|
| 832 |
-
|
|
|
|
| 595 |
def build_interface():
|
| 596 |
with gr.Blocks(title="Multimodal AI Demo", theme=gr.themes.Soft()) as demo_block:
|
| 597 |
gr.Markdown("# AI модели")
|
| 598 |
+
|
| 599 |
+
with gr.Tab("Классификация аудио"):
|
| 600 |
+
gr.Markdown("## Классификация аудио")
|
| 601 |
+
with gr.Row():
|
| 602 |
+
audio_input_component = gr.Audio(
|
| 603 |
+
label="Загрузите аудиофайл",
|
| 604 |
+
type="filepath",
|
| 605 |
+
)
|
| 606 |
+
audio_model_selector = gr.Dropdown(
|
| 607 |
+
choices=["audio_classifier", "emotion_classifier"],
|
| 608 |
+
label="Выберите модель",
|
| 609 |
+
value="audio_classifier",
|
| 610 |
+
info=(
|
| 611 |
+
"audio_classifier - общая классификация (курс)"
|
| 612 |
+
"emotion_classifier - эмоции в речи "
|
| 613 |
+
),
|
| 614 |
+
)
|
| 615 |
+
audio_classify_button = gr.Button("Применить")
|
| 616 |
+
|
| 617 |
+
audio_output_component = gr.Textbox(
|
| 618 |
+
label="Результаты классификации",
|
| 619 |
+
lines=10,
|
| 620 |
+
)
|
| 621 |
+
|
| 622 |
+
audio_classify_button.click(
|
| 623 |
+
fn=classify_audio_file,
|
| 624 |
+
inputs=[audio_input_component, audio_model_selector],
|
| 625 |
+
outputs=audio_output_component,
|
| 626 |
+
)
|
| 627 |
+
|
| 628 |
+
with gr.Tab("Zero-Shot аудио"):
|
| 629 |
+
gr.Markdown("## Zero-Shot аудио классификатор")
|
| 630 |
+
with gr.Row():
|
| 631 |
+
clap_audio_input_component = gr.Audio(
|
| 632 |
+
label="Загрузите аудиофайл",
|
| 633 |
+
type="filepath",
|
| 634 |
+
)
|
| 635 |
+
clap_label_texts_component = gr.Textbox(
|
| 636 |
+
label="Кандидатные метки (через запятую)",
|
| 637 |
+
placeholder="лай собаки, шум дождя, музыка, разговор",
|
| 638 |
+
lines=2,
|
| 639 |
+
)
|
| 640 |
+
clap_button = gr.Button("Применить")
|
| 641 |
+
|
| 642 |
+
clap_output_component = gr.Textbox(
|
| 643 |
+
label="Результаты zero-shot классификации",
|
| 644 |
+
lines=10,
|
| 645 |
+
)
|
| 646 |
+
|
| 647 |
+
clap_button.click(
|
| 648 |
+
fn=classify_audio_zero_shot_clap,
|
| 649 |
+
inputs=[clap_audio_input_component, clap_label_texts_component],
|
| 650 |
+
outputs=clap_output_component,
|
| 651 |
+
)
|
| 652 |
+
|
| 653 |
+
with gr.Tab("Распознавание речи"):
|
| 654 |
+
gr.Markdown("## Распознавание реч")
|
| 655 |
+
with gr.Row():
|
| 656 |
+
asr_audio_input_component = gr.Audio(
|
| 657 |
+
label="Загрузите аудио с речью",
|
| 658 |
+
type="filepath",
|
| 659 |
+
)
|
| 660 |
+
asr_model_selector = gr.Dropdown(
|
| 661 |
+
choices=["whisper", "wav2vec2"],
|
| 662 |
+
label="Выберите модель",
|
| 663 |
+
value="whisper",
|
| 664 |
+
info=(
|
| 665 |
+
"whisper - distil-whisper/distil-small.en (курс),\n"
|
| 666 |
+
"wav2vec2 - openai/whisper-small"
|
| 667 |
+
),
|
| 668 |
+
)
|
| 669 |
+
asr_button = gr.Button("Применить")
|
| 670 |
+
|
| 671 |
+
asr_output_component = gr.Textbox(
|
| 672 |
+
label="Транскрипция",
|
| 673 |
+
lines=5,
|
| 674 |
+
)
|
| 675 |
+
|
| 676 |
+
asr_button.click(
|
| 677 |
+
fn=recognize_speech,
|
| 678 |
+
inputs=[asr_audio_input_component, asr_model_selector],
|
| 679 |
+
outputs=asr_output_component,
|
| 680 |
+
)
|
| 681 |
+
with gr.Tab("Синтез речи"):
|
| 682 |
+
gr.Markdown("## Text-to-Speech")
|
| 683 |
+
with gr.Row():
|
| 684 |
+
tts_text_component = gr.Textbox(
|
| 685 |
+
label="Введите текст для синтеза",
|
| 686 |
+
placeholder="Введите текст на русском или английском языке...",
|
| 687 |
+
lines=3,
|
| 688 |
+
)
|
| 689 |
+
tts_model_selector = gr.Dropdown(
|
| 690 |
+
choices=["mms", "Google TTS"],
|
| 691 |
+
label="Выберите модель",
|
| 692 |
+
value="mms",
|
| 693 |
+
info=(
|
| 694 |
+
"facebook/mms-tts-rus\n"
|
| 695 |
+
"Google TTS"
|
| 696 |
+
),
|
| 697 |
+
)
|
| 698 |
+
tts_button = gr.Button("Применить")
|
| 699 |
+
|
| 700 |
+
tts_audio_output_component = gr.Audio(
|
| 701 |
+
label="Синтезированная речь",
|
| 702 |
+
type="filepath",
|
| 703 |
+
)
|
| 704 |
+
|
| 705 |
+
tts_button.click(
|
| 706 |
+
fn=synthesize_speech,
|
| 707 |
+
inputs=[tts_text_component, tts_model_selector],
|
| 708 |
+
outputs=tts_audio_output_component,
|
| 709 |
+
)
|
| 710 |
+
|
| 711 |
with gr.Tab("Детекция объектов"):
|
| 712 |
gr.Markdown("## Детекция объектов")
|
| 713 |
with gr.Row():
|
|
|
|
| 940 |
|
| 941 |
if __name__ == "__main__":
|
| 942 |
interface_block = build_interface()
|
| 943 |
+
interface_block.launch(share=True)
|
|
|