/j#ddlmZddlmZddlmZddlZddlmZddl m Z m Z m Z ddl mZddlmZmZd ZGd d eZGd d eZdS)) annotations)Path)AnyN) YOLODataset)ComposeFormat v8_transforms)DetectionValidator)colorstrops)RTDETRValidatorc<eZdZdZddfd Zdfd Zd dZxZS) RTDETRDatasetaReal-Time DEtection and TRacking (RT-DETR) dataset class extending the base YOLODataset class. This specialized dataset class is designed for use with the RT-DETR object detection model and is optimized for real-time detection and tracking tasks. Attributes: augment (bool): Whether to apply data augmentation. rect (bool): Whether to use rectangular training. use_segments (bool): Whether to use segmentation masks. use_keypoints (bool): Whether to use keypoint annotations. imgsz (int): Target image size for training. Methods: load_image: Load one image from dataset index. build_transforms: Build transformation pipeline for the dataset. Examples: Initialize an RT-DETR dataset >>> dataset = RTDETRDataset(img_path="path/to/images", imgsz=640) >>> image, hw0, hw = dataset.load_image(0) N)datac>tj|d|i|dS)aLInitialize the RTDETRDataset class by inheriting from the YOLODataset class. This constructor sets up a dataset specifically optimized for the RT-DETR (Real-Time DEtection and TRacking) model, building upon the base YOLODataset functionality. Args: *args (Any): Variable length argument list passed to the parent YOLODataset class. data (dict | None): Dictionary containing dataset information. If None, default values will be used. **kwargs (Any): Additional keyword arguments passed to the parent YOLODataset class. rN)super__init__)selfrargskwargs __class__s b/home/longshao/multi-rider-rag/.venv/lib/python3.11/site-packages/ultralytics/models/rtdetr/val.pyrzRTDETRDataset.__init__)s, $4T4V44444FcJt||S)aLoad one image from dataset index 'i'. Args: i (int): Index of the image to load. rect_mode (bool, optional): Whether to use rectangular mode for batch inference. Returns: im (np.ndarray): Loaded image as a NumPy array. hw_original (tuple[int, int]): Original image dimensions in (height, width) format. hw_resized (tuple[int, int]): Resized image dimensions in (height, width) format. Examples: Load an image from the dataset >>> dataset = RTDETRDataset(img_path="path/to/images") >>> image, hw0, hw = dataset.load_image(0) )i rect_mode)r load_image)rrrrs rrzRTDETRDataset.load_image6s""ww!!A!;;;rc |jrm|jr|js|jnd|_|jr|js|jnd|_|jr|js|jnd|_t ||j|d}ntg}|tdd|j |j d|j |j |S)zBuild transformation pipeline for the dataset. Args: hyp (dict, optional): Hyperparameters for transformations. Returns: (Compose): Composition of transformation functions. gT)stretchxywh) bbox_format normalize return_maskreturn_keypoint batch_idx mask_ratio mask_overlap)augmentrectmosaicmixupcutmixr imgszrappendr use_segments use_keypointsr& overlap_mask)rhyp transformss rbuild_transformszRTDETRDataset.build_transformsIs < %'+|NDIN3CJ%)\L$)L CI'+|NDIN3CJ&tTZdKKKJJ!J " - $ 2> -    r)F)N)__name__ __module__ __qualname____doc__rrr4 __classcell__)rs@rrrs,$( 5 5 5 5 5 5 5<<<<<<&rrc2eZdZdZddZdd ZddZddZdS)r aRTDETRValidator extends the DetectionValidator class to provide validation capabilities specifically tailored for the RT-DETR (Real-Time DETR) object detection model. The class allows building of an RTDETR-specific dataset for validation, applies confidence thresholding for post-processing, and updates evaluation metrics accordingly. Attributes: args (Namespace): Configuration arguments for validation. data (dict): Dataset configuration dictionary. Methods: build_dataset: Build an RTDETR Dataset for validation. postprocess: Apply confidence thresholding to prediction outputs. Examples: Initialize and run RT-DETR validation >>> from ultralytics.models.rtdetr import RTDETRValidator >>> args = dict(model="rtdetr-l.pt", data="coco8.yaml") >>> validator = RTDETRValidator(args=args) >>> validator() Notes: For further details on the attributes and methods, refer to the parent DetectionValidator class. valNc t||jj|d|jd|jjpdt |d|j S)aBuild an RTDETR Dataset. Args: img_path (str): Path to the folder containing images. mode (str, optional): `train` mode or `val` mode, users are able to customize different augmentations for each mode. batch (int, optional): Size of batches, this is for `rect`. Returns: (RTDETRDataset): Dataset configured for RT-DETR validation. FNz: ) img_pathr- batch_sizer(r2r)cacheprefixr)rrr-r?r r)rr=modebatchs r build_datasetzRTDETRValidator.build_datasetsU)/ )/)TtKKK((    rpredndict[str, torch.Tensor]pbatchdict[str, Any]returnc|S)zJReturn predictions unchanged as RT-DETR handles scaling in postprocessing.)rrDrFs r scale_predszRTDETRValidator.scale_predss rpreds7torch.Tensor | list[torch.Tensor] | tuple[torch.Tensor]list[dict[str, torch.Tensor]]cdt|ttfs|dg}|dj\}}}|dd|dz fd\}}||jjz}tjd|j g|z}t|D]\}} tj | } || d\} } tj| | d| dgd} | | d } | | |jjk||<d |DS) aApply confidence thresholding to prediction outputs. Args: preds (torch.Tensor | list | tuple): Raw predictions from the model. If tensor, should have shape (batch_size, num_predictions, num_classes + 4) where last dimension contains bbox coords and class scores. Returns: (list[dict[str, torch.Tensor]]): List of dictionaries for each image, each containing: - 'bboxes': Tensor of shape (N, 4) with bounding box coordinates - 'conf': Tensor of shape (N,) with confidence scores - 'cls': Tensor of shape (N,) with class indices Nr)dim)r)device).NT) descendingcZg|](}|ddddf|dddf|dddfd)S)NrP)bboxesconfclsrJ.0xs r z/RTDETRValidator.postprocess..sJWWW!1QQQU8Qqqq!tWQqqq!tWEEWWWr) isinstancelisttupleshapesplitrr-torchzerosrT enumerater xywh2xyxymaxcatargsortrY) rrLbs_ndrXscoresoutputsrbboxscorerZpreds r postprocesszRTDETRValidator.postprocesss1 %$// "DME!HN ArqBF <<$)/!;vfm<<<=B (( 6 6GAt=&&Dr**JE39dE)$4c)nE2NNND  667Dedin45GAJJWWwWWWWrNonec @t|d}|j}|rt|n|}|d}|dddgfxx|dd|jjz zcc<|dddgfxx|dd|jjz zcc<tj|}|d d d dfxx|d d dd fdz zcc<t| |d  |d  D][\}}} |j ||j |jt| d |Dt|d d\d S)aSerialize YOLO predictions to COCO json format. Args: predn (dict[str, torch.Tensor]): Predictions dictionary containing 'bboxes', 'conf', and 'cls' keys with bounding box coordinates, confidence scores, and class predictions. pbatch (dict[str, Any]): Batch dictionary containing 'imgsz', 'ori_shape', 'ratio_pad', and 'im_file'. im_filerX.r ori_shapeNrYrZc.g|]}t|dS)rz)roundr[s rr^z0RTDETRValidator.pred_to_json..s 444QU1a[[444rrW)image_id file_name category_idrprq)rstem isnumericintclonerr-r xyxy2xywhziptolistjdictr.name class_mapr|) rrDrFpathrr}boxbscs r pred_to_jsonzRTDETRValidator.pred_to_jsonsF9%&&y $ 0 0:3t999dHo##%% C!QKF;/2TY_DD C!QKF;/2TY_DDmC   AAArrE c!!!QRR%j1n$ 3::<<v)=)=)?)?uATATAVAVWW  GAq! J   (!%#'>#a& 44!444"1a[[       r)r;N)rDrErFrGrHrE)rLrMrHrN)rDrErFrGrHrt)r5r6r7r8rCrKrsrrJrrr r hss2    0XXXXBrr ) __future__rpathlibrtypingrrdultralytics.datarultralytics.data.augmentrrr ultralytics.models.yolo.detectr ultralytics.utilsr r __all__rr rJrrrs#""""" ((((((CCCCCCCCCC======++++++++ SSSSSKSSSlppppp(pppppr