import gradio as gr
import cv2
import insightface
from insightface.app import FaceAnalysis

def predict(image_in_video, image_in_img):
    if image_in_video == None and image_in_img == None:
        raise gr.Error("Please capture an image using the webcam or upload an image.")
    image = image_in_video or image_in_img
    return swapi(image)

app = FaceAnalysis(name='buffalo_l')
app.prepare(ctx_id=0, det_size=(640, 640))
swapper = insightface.model_zoo.get_model('inswapper_128.onnx', download='FALSE', download_zip= 'FALSE')

def swapi(imagen):
    img = cv2.cvtColor(np.array(imagen), cv2.COLOR_RGB2BGR)  # Convert image from RGB to BGR format
    faces = app.get(img)

    if not faces:
        return img  # If no faces are detected, return the original image

    source_face = faces[0]
    bbox = source_face['bbox']
    bbox = [int(b) for b in bbox]

    res = img.copy()
    for face in faces:
        res = swapper.get(res, face, source_face, paste_back=True)

    return res[:, :, [2, 1, 0]]  # Convert BGR to RGB for Gradio display

with gr.Blocks() as blocks:
    gr.Markdown("### Capture Image Using WebCam or Upload") 

    with gr.Row():
        with gr.Column():
            image_or_file_opt = gr.Radio(["webcam", "file"], value="webcam",
                                         label="How would you like to upload your image?")
            image_in_video = gr.Image(source="webcam", type="filepath")
            image_in_img = gr.Image(source="upload", visible=False, type="filepath")

            # Update visibility based on selection
            def toggle(choice):
                if choice == "webcam":
                    return gr.update(visible=True, value=None), gr.update(visible=False, value=None)
                else:
                    return gr.update(visible=False, value=None), gr.update(visible=True, value=None)

            image_or_file_opt.change(fn=toggle, inputs=[image_or_file_opt],
                                     outputs=[image_in_video, image_in_img], queue=False, show_progress=False)
        with gr.Column():
            image_out = gr.Image()

    run_btn = gr.Button("Run")
    run_btn.click(fn=predict, inputs=[image_in_img, image_in_video], outputs=[image_out])
    gr.Examples(fn=predict, examples=[], inputs=[image_in_img, image_in_video], outputs=[image_out])

blocks.queue()
blocks.launch()