app.py

import gradio as gr
from ultralytics import YOLO
import speech_recognition as sr

model = YOLO('best.pt')

def proses_image(image):
    results = model(image)
    return results[0].plot()

def process_audio(audio):
    recognizer = sr.Recognizer()
    with sr.AudioFile(audio) as source:
        audio_data = recognizer.record(source)
        try:
            text = recognizer.recognize_google(audio_data)
            return text
        except sr.UnknownValueError:
            return "Audio tidak dapat dikenali."
        except sr.RequestError as e:
            return f"Error dengan layanan pengenalan suara: {e}"

def update_visibility(input):
    if input == "Gambar":
        return gr.Row(visible=True), gr.Row(visible=False)
    elif input == "Audio":
        return gr.Row(visible=False), gr.Row(visible=True)

with gr.Blocks() as demo:
    with gr.Row():
        gr.Markdown("# Multimodal America Sign Language")
    with gr.Row():
        input = gr.Radio(["Gambar", "Audio"], value="Gambar", label="Pilih mode:")
    with gr.Row(visible=True) as gambar:
        gr.Interface(
            fn=proses_image,
            inputs=gr.Image(),
            outputs=gr.Image(),
            live=True
        )
    with gr.Row(visible=False) as audio:
        gr.Interface(
            fn=process_audio,
            inputs=gr.Audio(sources="microphone", type="filepath", streaming=True),
            outputs=gr.Textbox(),
            live=True
        )

    input.change(update_visibility, inputs=[input], outputs=[gambar, audio])

demo.launch()