kadirnar/Yolov10

1import gradio as gr2from ultralytics import YOLO3import supervision as sv4 5    6box_annotator = sv.BoxAnnotator()7category_dict = {8    0: 'person', 1: 'bicycle', 2: 'car', 3: 'motorcycle', 4: 'airplane', 5: 'bus',9    6: 'train', 7: 'truck', 8: 'boat', 9: 'traffic light', 10: 'fire hydrant',10    11: 'stop sign', 12: 'parking meter', 13: 'bench', 14: 'bird', 15: 'cat',11    16: 'dog', 17: 'horse', 18: 'sheep', 19: 'cow', 20: 'elephant', 21: 'bear',12    22: 'zebra', 23: 'giraffe', 24: 'backpack', 25: 'umbrella', 26: 'handbag',13    27: 'tie', 28: 'suitcase', 29: 'frisbee', 30: 'skis', 31: 'snowboard',14    32: 'sports ball', 33: 'kite', 34: 'baseball bat', 35: 'baseball glove',15    36: 'skateboard', 37: 'surfboard', 38: 'tennis racket', 39: 'bottle',16    40: 'wine glass', 41: 'cup', 42: 'fork', 43: 'knife', 44: 'spoon', 45: 'bowl',17    46: 'banana', 47: 'apple', 48: 'sandwich', 49: 'orange', 50: 'broccoli',18    51: 'carrot', 52: 'hot dog', 53: 'pizza', 54: 'donut', 55: 'cake',19    56: 'chair', 57: 'couch', 58: 'potted plant', 59: 'bed', 60: 'dining table',20    61: 'toilet', 62: 'tv', 63: 'laptop', 64: 'mouse', 65: 'remote', 66: 'keyboard',21    67: 'cell phone', 68: 'microwave', 69: 'oven', 70: 'toaster', 71: 'sink',22    72: 'refrigerator', 73: 'book', 74: 'clock', 75: 'vase', 76: 'scissors',23    77: 'teddy bear', 78: 'hair drier', 79: 'toothbrush'24}25 26 27 28def yolov10_inference(image, model_id, image_size, conf_threshold, iou_threshold):29    model = YOLO(f"{model_id}.pt")30    results = model(source=image, imgsz=image_size, iou=iou_threshold, conf=conf_threshold, verbose=False)[0]31    detections = sv.Detections.from_ultralytics(results)32    labels = [33        f"{category_dict[class_id]} {confidence:.2f}"34        for class_id, confidence in zip(detections.class_id, detections.confidence)35    ]36    annotated_image = box_annotator.annotate(image, detections=detections, labels=labels)37 38    return annotated_image39 40def app():41    with gr.Blocks():42        with gr.Row():43            with gr.Column():44                image = gr.Image(type="pil", label="Image")45                46                model_id = gr.Dropdown(47                    label="Model",48                    choices=[49                        "yolov10n",50                        "yolov10s",51                        "yolov10m",52                        "yolov10b",53                        "yolov10l",54                        "yolov10x",55                    ],56                    value="yolov10m",57                )58                image_size = gr.Slider(59                    label="Image Size",60                    minimum=320,61                    maximum=1280,62                    step=32,63                    value=640,64                )65                conf_threshold = gr.Slider(66                    label="Confidence Threshold",67                    minimum=0.1,68                    maximum=1.0,69                    step=0.1,70                    value=0.25,71                )72                iou_threshold = gr.Slider(73                    label="IoU Threshold",74                    minimum=0.1,75                    maximum=1.0,76                    step=0.1,77                    value=0.45,78                )79                yolov10_infer = gr.Button(value="Detect Objects")80 81            with gr.Column():82                output_image = gr.Image(type="pil", label="Annotated Image")83 84        yolov10_infer.click(85            fn=yolov10_inference,86            inputs=[87                image,88                model_id,89                image_size,90                conf_threshold,91                iou_threshold,92            ],93            outputs=[output_image],94        )95 96        gr.Examples(97            examples=[98                [99                    "dog.jpeg",100                    "yolov10x",101                    640,102                    0.25,103                    0.45,104                ],105                [106                    "huggingface.jpg",107                    "yolov10m",108                    640,109                    0.25,110                    0.45,111                ],112                [113                    "zidane.jpg",114                    "yolov10b",115                    640,116                    0.25,117                    0.45,118                ],119            ],120            fn=yolov10_inference,121            inputs=[122                image,123                model_id,124                image_size,125                conf_threshold,126                iou_threshold,127            ],128            outputs=[output_image],129            cache_examples="lazy",130        )131 132gradio_app = gr.Blocks()133with gradio_app:134    gr.HTML(135        """136    <h1 style='text-align: center'>137    YOLOv10: Real-Time End-to-End Object Detection138    </h1>139    """)140    gr.HTML(141        """142        <h3 style='text-align: center'>143        Follow me for more!144        <a href='https://twitter.com/kadirnar_ai' target='_blank'>Twitter</a> | <a href='https://github.com/kadirnar' target='_blank'>Github</a> | <a href='https://www.linkedin.com/in/kadir-nar/' target='_blank'>Linkedin</a>  | <a href='https://www.huggingface.co/kadirnar/' target='_blank'>HuggingFace</a>145        </h3>146        """)147    with gr.Row():148        with gr.Column():149            app()150 151gradio_app.launch(debug=True)