import spaces
import gradio as gr
import cv2
import tempfile
from ultralytics import YOLOv10

image_processor = RTDetrImageProcessor.from_pretrained("PekingU/rtdetr_r50vd")
model = RTDetrForObjectDetection.from_pretrained("PekingU/rtdetr_r50vd")

@spaces.GPU
def yolov10_inference(image, conf_threshold):

    inputs = image_processor(images=image, return_tensors="pt")

    with torch.no_grad():
        outputs = model(**inputs)

    results = image_processor.post_process_object_detection(outputs, target_sizes=torch.tensor([image.size[::-1]]), threshold=0.3)



def app():
    with gr.Blocks():
        with gr.Row():
            with gr.Column():
                image = gr.Image(type="pil", label="Image", visible=True, sources="webcam", height=500, width=500)
                conf_threshold = gr.Slider(
                    label="Confidence Threshold",
                    minimum=0.0,
                    maximum=1.0,
                    step=0.05,
                    value=0.25,
                )
        image.stream(
            fn=yolov10_inference,
            inputs=[image, conf_threshold],
            outputs=[image],
            stream_every=0.2,
            time_limit=30
        )


gradio_app = gr.Blocks()
with gradio_app:
    gr.HTML(
        """
    <h1 style='text-align: center'>
    YOLOv10 Webcam Stream
    </h1>
    """)
    gr.HTML(
        """
        <h3 style='text-align: center'>
        <a href='https://arxiv.org/abs/2405.14458' target='_blank'>arXiv</a> | <a href='https://github.com/THU-MIG/yolov10' target='_blank'>github</a>
        </h3>
        """)
    with gr.Row():
        with gr.Column():
            app()
if __name__ == '__main__':
    gradio_app.launch()