Salta para o conteúdo

Referência para ultralytics/utils/tal.py

Nota

Este ficheiro está disponível em https://github.com/ultralytics/ ultralytics/blob/main/ ultralytics/utils/tal .py. Se detectares um problema, por favor ajuda a corrigi-lo contribuindo com um Pull Request 🛠️. Obrigado 🙏!



ultralytics.utils.tal.TaskAlignedAssigner

Bases: Module

Um atribuidor alinhado com a tarefa para a deteção de objectos.

Esta classe atribui objectos da verdade terrestre (gt) a âncoras com base na métrica alinhada com a tarefa, que combina classificação e informação de localização.

Atributos:

Nome Tipo Descrição
topk int

O número de candidatos de topo a considerar.

num_classes int

O número de classes de objectos.

alpha float

O parâmetro alfa para a componente de classificação da métrica de alinhamento de tarefas.

beta float

O parâmetro beta para a componente de localização da métrica alinhada com a tarefa.

eps float

Um valor pequeno para evitar a divisão por zero.

Código fonte em ultralytics/utils/tal.py
class TaskAlignedAssigner(nn.Module):
    """
    A task-aligned assigner for object detection.

    This class assigns ground-truth (gt) objects to anchors based on the task-aligned metric, which combines both
    classification and localization information.

    Attributes:
        topk (int): The number of top candidates to consider.
        num_classes (int): The number of object classes.
        alpha (float): The alpha parameter for the classification component of the task-aligned metric.
        beta (float): The beta parameter for the localization component of the task-aligned metric.
        eps (float): A small value to prevent division by zero.
    """

    def __init__(self, topk=13, num_classes=80, alpha=1.0, beta=6.0, eps=1e-9):
        """Initialize a TaskAlignedAssigner object with customizable hyperparameters."""
        super().__init__()
        self.topk = topk
        self.num_classes = num_classes
        self.bg_idx = num_classes
        self.alpha = alpha
        self.beta = beta
        self.eps = eps

    @torch.no_grad()
    def forward(self, pd_scores, pd_bboxes, anc_points, gt_labels, gt_bboxes, mask_gt):
        """
        Compute the task-aligned assignment. Reference code is available at
        https://github.com/Nioolek/PPYOLOE_pytorch/blob/master/ppyoloe/assigner/tal_assigner.py.

        Args:
            pd_scores (Tensor): shape(bs, num_total_anchors, num_classes)
            pd_bboxes (Tensor): shape(bs, num_total_anchors, 4)
            anc_points (Tensor): shape(num_total_anchors, 2)
            gt_labels (Tensor): shape(bs, n_max_boxes, 1)
            gt_bboxes (Tensor): shape(bs, n_max_boxes, 4)
            mask_gt (Tensor): shape(bs, n_max_boxes, 1)

        Returns:
            target_labels (Tensor): shape(bs, num_total_anchors)
            target_bboxes (Tensor): shape(bs, num_total_anchors, 4)
            target_scores (Tensor): shape(bs, num_total_anchors, num_classes)
            fg_mask (Tensor): shape(bs, num_total_anchors)
            target_gt_idx (Tensor): shape(bs, num_total_anchors)
        """
        self.bs = pd_scores.shape[0]
        self.n_max_boxes = gt_bboxes.shape[1]

        if self.n_max_boxes == 0:
            device = gt_bboxes.device
            return (
                torch.full_like(pd_scores[..., 0], self.bg_idx).to(device),
                torch.zeros_like(pd_bboxes).to(device),
                torch.zeros_like(pd_scores).to(device),
                torch.zeros_like(pd_scores[..., 0]).to(device),
                torch.zeros_like(pd_scores[..., 0]).to(device),
            )

        mask_pos, align_metric, overlaps = self.get_pos_mask(
            pd_scores, pd_bboxes, gt_labels, gt_bboxes, anc_points, mask_gt
        )

        target_gt_idx, fg_mask, mask_pos = self.select_highest_overlaps(mask_pos, overlaps, self.n_max_boxes)

        # Assigned target
        target_labels, target_bboxes, target_scores = self.get_targets(gt_labels, gt_bboxes, target_gt_idx, fg_mask)

        # Normalize
        align_metric *= mask_pos
        pos_align_metrics = align_metric.amax(dim=-1, keepdim=True)  # b, max_num_obj
        pos_overlaps = (overlaps * mask_pos).amax(dim=-1, keepdim=True)  # b, max_num_obj
        norm_align_metric = (align_metric * pos_overlaps / (pos_align_metrics + self.eps)).amax(-2).unsqueeze(-1)
        target_scores = target_scores * norm_align_metric

        return target_labels, target_bboxes, target_scores, fg_mask.bool(), target_gt_idx

    def get_pos_mask(self, pd_scores, pd_bboxes, gt_labels, gt_bboxes, anc_points, mask_gt):
        """Get in_gts mask, (b, max_num_obj, h*w)."""
        mask_in_gts = self.select_candidates_in_gts(anc_points, gt_bboxes)
        # Get anchor_align metric, (b, max_num_obj, h*w)
        align_metric, overlaps = self.get_box_metrics(pd_scores, pd_bboxes, gt_labels, gt_bboxes, mask_in_gts * mask_gt)
        # Get topk_metric mask, (b, max_num_obj, h*w)
        mask_topk = self.select_topk_candidates(align_metric, topk_mask=mask_gt.expand(-1, -1, self.topk).bool())
        # Merge all mask to a final mask, (b, max_num_obj, h*w)
        mask_pos = mask_topk * mask_in_gts * mask_gt

        return mask_pos, align_metric, overlaps

    def get_box_metrics(self, pd_scores, pd_bboxes, gt_labels, gt_bboxes, mask_gt):
        """Compute alignment metric given predicted and ground truth bounding boxes."""
        na = pd_bboxes.shape[-2]
        mask_gt = mask_gt.bool()  # b, max_num_obj, h*w
        overlaps = torch.zeros([self.bs, self.n_max_boxes, na], dtype=pd_bboxes.dtype, device=pd_bboxes.device)
        bbox_scores = torch.zeros([self.bs, self.n_max_boxes, na], dtype=pd_scores.dtype, device=pd_scores.device)

        ind = torch.zeros([2, self.bs, self.n_max_boxes], dtype=torch.long)  # 2, b, max_num_obj
        ind[0] = torch.arange(end=self.bs).view(-1, 1).expand(-1, self.n_max_boxes)  # b, max_num_obj
        ind[1] = gt_labels.squeeze(-1)  # b, max_num_obj
        # Get the scores of each grid for each gt cls
        bbox_scores[mask_gt] = pd_scores[ind[0], :, ind[1]][mask_gt]  # b, max_num_obj, h*w

        # (b, max_num_obj, 1, 4), (b, 1, h*w, 4)
        pd_boxes = pd_bboxes.unsqueeze(1).expand(-1, self.n_max_boxes, -1, -1)[mask_gt]
        gt_boxes = gt_bboxes.unsqueeze(2).expand(-1, -1, na, -1)[mask_gt]
        overlaps[mask_gt] = self.iou_calculation(gt_boxes, pd_boxes)

        align_metric = bbox_scores.pow(self.alpha) * overlaps.pow(self.beta)
        return align_metric, overlaps

    def iou_calculation(self, gt_bboxes, pd_bboxes):
        """IoU calculation for horizontal bounding boxes."""
        return bbox_iou(gt_bboxes, pd_bboxes, xywh=False, CIoU=True).squeeze(-1).clamp_(0)

    def select_topk_candidates(self, metrics, largest=True, topk_mask=None):
        """
        Select the top-k candidates based on the given metrics.

        Args:
            metrics (Tensor): A tensor of shape (b, max_num_obj, h*w), where b is the batch size,
                              max_num_obj is the maximum number of objects, and h*w represents the
                              total number of anchor points.
            largest (bool): If True, select the largest values; otherwise, select the smallest values.
            topk_mask (Tensor): An optional boolean tensor of shape (b, max_num_obj, topk), where
                                topk is the number of top candidates to consider. If not provided,
                                the top-k values are automatically computed based on the given metrics.

        Returns:
            (Tensor): A tensor of shape (b, max_num_obj, h*w) containing the selected top-k candidates.
        """

        # (b, max_num_obj, topk)
        topk_metrics, topk_idxs = torch.topk(metrics, self.topk, dim=-1, largest=largest)
        if topk_mask is None:
            topk_mask = (topk_metrics.max(-1, keepdim=True)[0] > self.eps).expand_as(topk_idxs)
        # (b, max_num_obj, topk)
        topk_idxs.masked_fill_(~topk_mask, 0)

        # (b, max_num_obj, topk, h*w) -> (b, max_num_obj, h*w)
        count_tensor = torch.zeros(metrics.shape, dtype=torch.int8, device=topk_idxs.device)
        ones = torch.ones_like(topk_idxs[:, :, :1], dtype=torch.int8, device=topk_idxs.device)
        for k in range(self.topk):
            # Expand topk_idxs for each value of k and add 1 at the specified positions
            count_tensor.scatter_add_(-1, topk_idxs[:, :, k : k + 1], ones)
        # count_tensor.scatter_add_(-1, topk_idxs, torch.ones_like(topk_idxs, dtype=torch.int8, device=topk_idxs.device))
        # Filter invalid bboxes
        count_tensor.masked_fill_(count_tensor > 1, 0)

        return count_tensor.to(metrics.dtype)

    def get_targets(self, gt_labels, gt_bboxes, target_gt_idx, fg_mask):
        """
        Compute target labels, target bounding boxes, and target scores for the positive anchor points.

        Args:
            gt_labels (Tensor): Ground truth labels of shape (b, max_num_obj, 1), where b is the
                                batch size and max_num_obj is the maximum number of objects.
            gt_bboxes (Tensor): Ground truth bounding boxes of shape (b, max_num_obj, 4).
            target_gt_idx (Tensor): Indices of the assigned ground truth objects for positive
                                    anchor points, with shape (b, h*w), where h*w is the total
                                    number of anchor points.
            fg_mask (Tensor): A boolean tensor of shape (b, h*w) indicating the positive
                              (foreground) anchor points.

        Returns:
            (Tuple[Tensor, Tensor, Tensor]): A tuple containing the following tensors:
                - target_labels (Tensor): Shape (b, h*w), containing the target labels for
                                          positive anchor points.
                - target_bboxes (Tensor): Shape (b, h*w, 4), containing the target bounding boxes
                                          for positive anchor points.
                - target_scores (Tensor): Shape (b, h*w, num_classes), containing the target scores
                                          for positive anchor points, where num_classes is the number
                                          of object classes.
        """

        # Assigned target labels, (b, 1)
        batch_ind = torch.arange(end=self.bs, dtype=torch.int64, device=gt_labels.device)[..., None]
        target_gt_idx = target_gt_idx + batch_ind * self.n_max_boxes  # (b, h*w)
        target_labels = gt_labels.long().flatten()[target_gt_idx]  # (b, h*w)

        # Assigned target boxes, (b, max_num_obj, 4) -> (b, h*w, 4)
        target_bboxes = gt_bboxes.view(-1, gt_bboxes.shape[-1])[target_gt_idx]

        # Assigned target scores
        target_labels.clamp_(0)

        # 10x faster than F.one_hot()
        target_scores = torch.zeros(
            (target_labels.shape[0], target_labels.shape[1], self.num_classes),
            dtype=torch.int64,
            device=target_labels.device,
        )  # (b, h*w, 80)
        target_scores.scatter_(2, target_labels.unsqueeze(-1), 1)

        fg_scores_mask = fg_mask[:, :, None].repeat(1, 1, self.num_classes)  # (b, h*w, 80)
        target_scores = torch.where(fg_scores_mask > 0, target_scores, 0)

        return target_labels, target_bboxes, target_scores

    @staticmethod
    def select_candidates_in_gts(xy_centers, gt_bboxes, eps=1e-9):
        """
        Select the positive anchor center in gt.

        Args:
            xy_centers (Tensor): shape(h*w, 2)
            gt_bboxes (Tensor): shape(b, n_boxes, 4)

        Returns:
            (Tensor): shape(b, n_boxes, h*w)
        """
        n_anchors = xy_centers.shape[0]
        bs, n_boxes, _ = gt_bboxes.shape
        lt, rb = gt_bboxes.view(-1, 1, 4).chunk(2, 2)  # left-top, right-bottom
        bbox_deltas = torch.cat((xy_centers[None] - lt, rb - xy_centers[None]), dim=2).view(bs, n_boxes, n_anchors, -1)
        # return (bbox_deltas.min(3)[0] > eps).to(gt_bboxes.dtype)
        return bbox_deltas.amin(3).gt_(eps)

    @staticmethod
    def select_highest_overlaps(mask_pos, overlaps, n_max_boxes):
        """
        If an anchor box is assigned to multiple gts, the one with the highest IoU will be selected.

        Args:
            mask_pos (Tensor): shape(b, n_max_boxes, h*w)
            overlaps (Tensor): shape(b, n_max_boxes, h*w)

        Returns:
            target_gt_idx (Tensor): shape(b, h*w)
            fg_mask (Tensor): shape(b, h*w)
            mask_pos (Tensor): shape(b, n_max_boxes, h*w)
        """
        # (b, n_max_boxes, h*w) -> (b, h*w)
        fg_mask = mask_pos.sum(-2)
        if fg_mask.max() > 1:  # one anchor is assigned to multiple gt_bboxes
            mask_multi_gts = (fg_mask.unsqueeze(1) > 1).expand(-1, n_max_boxes, -1)  # (b, n_max_boxes, h*w)
            max_overlaps_idx = overlaps.argmax(1)  # (b, h*w)

            is_max_overlaps = torch.zeros(mask_pos.shape, dtype=mask_pos.dtype, device=mask_pos.device)
            is_max_overlaps.scatter_(1, max_overlaps_idx.unsqueeze(1), 1)

            mask_pos = torch.where(mask_multi_gts, is_max_overlaps, mask_pos).float()  # (b, n_max_boxes, h*w)
            fg_mask = mask_pos.sum(-2)
        # Find each grid serve which gt(index)
        target_gt_idx = mask_pos.argmax(-2)  # (b, h*w)
        return target_gt_idx, fg_mask, mask_pos

__init__(topk=13, num_classes=80, alpha=1.0, beta=6.0, eps=1e-09)

Inicializa um objeto TaskAlignedAssigner com hiperparâmetros personalizáveis.

Código fonte em ultralytics/utils/tal.py
def __init__(self, topk=13, num_classes=80, alpha=1.0, beta=6.0, eps=1e-9):
    """Initialize a TaskAlignedAssigner object with customizable hyperparameters."""
    super().__init__()
    self.topk = topk
    self.num_classes = num_classes
    self.bg_idx = num_classes
    self.alpha = alpha
    self.beta = beta
    self.eps = eps

forward(pd_scores, pd_bboxes, anc_points, gt_labels, gt_bboxes, mask_gt)

Calcula a atribuição de tarefas alinhadas. O código de referência está disponível em https://github.com/Nioolek/PPYOLOE_pytorch/blob/master/ppyoloe/assigner/tal_assigner.py.

Parâmetros:

Nome Tipo Descrição Predefinição
pd_scores Tensor

shape(bs, num_total_anchors, num_classes)

necessário
pd_bboxes Tensor

shape(bs, num_total_anchors, 4)

necessário
anc_points Tensor

forma(num_total_âncoras, 2)

necessário
gt_labels Tensor

forma(bs, n_max_boxes, 1)

necessário
gt_bboxes Tensor

forma(bs, n_max_boxes, 4)

necessário
mask_gt Tensor

forma(bs, n_max_boxes, 1)

necessário

Devolve:

Nome Tipo Descrição
target_labels Tensor

shape(bs, num_total_anchors)

target_bboxes Tensor

shape(bs, num_total_anchors, 4)

target_scores Tensor

shape(bs, num_total_anchors, num_classes)

fg_mask Tensor

shape(bs, num_total_anchors)

target_gt_idx Tensor

shape(bs, num_total_anchors)

Código fonte em ultralytics/utils/tal.py
@torch.no_grad()
def forward(self, pd_scores, pd_bboxes, anc_points, gt_labels, gt_bboxes, mask_gt):
    """
    Compute the task-aligned assignment. Reference code is available at
    https://github.com/Nioolek/PPYOLOE_pytorch/blob/master/ppyoloe/assigner/tal_assigner.py.

    Args:
        pd_scores (Tensor): shape(bs, num_total_anchors, num_classes)
        pd_bboxes (Tensor): shape(bs, num_total_anchors, 4)
        anc_points (Tensor): shape(num_total_anchors, 2)
        gt_labels (Tensor): shape(bs, n_max_boxes, 1)
        gt_bboxes (Tensor): shape(bs, n_max_boxes, 4)
        mask_gt (Tensor): shape(bs, n_max_boxes, 1)

    Returns:
        target_labels (Tensor): shape(bs, num_total_anchors)
        target_bboxes (Tensor): shape(bs, num_total_anchors, 4)
        target_scores (Tensor): shape(bs, num_total_anchors, num_classes)
        fg_mask (Tensor): shape(bs, num_total_anchors)
        target_gt_idx (Tensor): shape(bs, num_total_anchors)
    """
    self.bs = pd_scores.shape[0]
    self.n_max_boxes = gt_bboxes.shape[1]

    if self.n_max_boxes == 0:
        device = gt_bboxes.device
        return (
            torch.full_like(pd_scores[..., 0], self.bg_idx).to(device),
            torch.zeros_like(pd_bboxes).to(device),
            torch.zeros_like(pd_scores).to(device),
            torch.zeros_like(pd_scores[..., 0]).to(device),
            torch.zeros_like(pd_scores[..., 0]).to(device),
        )

    mask_pos, align_metric, overlaps = self.get_pos_mask(
        pd_scores, pd_bboxes, gt_labels, gt_bboxes, anc_points, mask_gt
    )

    target_gt_idx, fg_mask, mask_pos = self.select_highest_overlaps(mask_pos, overlaps, self.n_max_boxes)

    # Assigned target
    target_labels, target_bboxes, target_scores = self.get_targets(gt_labels, gt_bboxes, target_gt_idx, fg_mask)

    # Normalize
    align_metric *= mask_pos
    pos_align_metrics = align_metric.amax(dim=-1, keepdim=True)  # b, max_num_obj
    pos_overlaps = (overlaps * mask_pos).amax(dim=-1, keepdim=True)  # b, max_num_obj
    norm_align_metric = (align_metric * pos_overlaps / (pos_align_metrics + self.eps)).amax(-2).unsqueeze(-1)
    target_scores = target_scores * norm_align_metric

    return target_labels, target_bboxes, target_scores, fg_mask.bool(), target_gt_idx

get_box_metrics(pd_scores, pd_bboxes, gt_labels, gt_bboxes, mask_gt)

Calcula a métrica de alinhamento com base nas caixas delimitadoras previstas e verdadeiras.

Código fonte em ultralytics/utils/tal.py
def get_box_metrics(self, pd_scores, pd_bboxes, gt_labels, gt_bboxes, mask_gt):
    """Compute alignment metric given predicted and ground truth bounding boxes."""
    na = pd_bboxes.shape[-2]
    mask_gt = mask_gt.bool()  # b, max_num_obj, h*w
    overlaps = torch.zeros([self.bs, self.n_max_boxes, na], dtype=pd_bboxes.dtype, device=pd_bboxes.device)
    bbox_scores = torch.zeros([self.bs, self.n_max_boxes, na], dtype=pd_scores.dtype, device=pd_scores.device)

    ind = torch.zeros([2, self.bs, self.n_max_boxes], dtype=torch.long)  # 2, b, max_num_obj
    ind[0] = torch.arange(end=self.bs).view(-1, 1).expand(-1, self.n_max_boxes)  # b, max_num_obj
    ind[1] = gt_labels.squeeze(-1)  # b, max_num_obj
    # Get the scores of each grid for each gt cls
    bbox_scores[mask_gt] = pd_scores[ind[0], :, ind[1]][mask_gt]  # b, max_num_obj, h*w

    # (b, max_num_obj, 1, 4), (b, 1, h*w, 4)
    pd_boxes = pd_bboxes.unsqueeze(1).expand(-1, self.n_max_boxes, -1, -1)[mask_gt]
    gt_boxes = gt_bboxes.unsqueeze(2).expand(-1, -1, na, -1)[mask_gt]
    overlaps[mask_gt] = self.iou_calculation(gt_boxes, pd_boxes)

    align_metric = bbox_scores.pow(self.alpha) * overlaps.pow(self.beta)
    return align_metric, overlaps

get_pos_mask(pd_scores, pd_bboxes, gt_labels, gt_bboxes, anc_points, mask_gt)

Obtém a máscara in_gts, (b, max_num_obj, h*w).

Código fonte em ultralytics/utils/tal.py
def get_pos_mask(self, pd_scores, pd_bboxes, gt_labels, gt_bboxes, anc_points, mask_gt):
    """Get in_gts mask, (b, max_num_obj, h*w)."""
    mask_in_gts = self.select_candidates_in_gts(anc_points, gt_bboxes)
    # Get anchor_align metric, (b, max_num_obj, h*w)
    align_metric, overlaps = self.get_box_metrics(pd_scores, pd_bboxes, gt_labels, gt_bboxes, mask_in_gts * mask_gt)
    # Get topk_metric mask, (b, max_num_obj, h*w)
    mask_topk = self.select_topk_candidates(align_metric, topk_mask=mask_gt.expand(-1, -1, self.topk).bool())
    # Merge all mask to a final mask, (b, max_num_obj, h*w)
    mask_pos = mask_topk * mask_in_gts * mask_gt

    return mask_pos, align_metric, overlaps

get_targets(gt_labels, gt_bboxes, target_gt_idx, fg_mask)

Calcula as etiquetas dos alvos, as caixas delimitadoras dos alvos e as pontuações dos alvos para os pontos de ancoragem positivos.

Parâmetros:

Nome Tipo Descrição Predefinição
gt_labels Tensor

Etiquetas de verdade terrestre da forma (b, max_num_obj, 1), em que b é o tamanho do lote e max_num_obj é o número máximo de objectos. tamanho do lote e max_num_obj é o número máximo de objectos.

necessário
gt_bboxes Tensor

Caixas delimitadoras da verdade terrestre da forma (b, max_num_obj, 4).

necessário
target_gt_idx Tensor

Índices dos objectos de verdade atribuídos para pontos de ancoragem pontos de ancoragem positivos, com a forma (b, hw), em que hwé o número total número total de pontos de ancoragem.

necessário
fg_mask Tensor

Um booleano tensor de forma (b, h*w) que indica os pontos de ancoragem positivos (primeiro plano). (primeiro plano) pontos de ancoragem positivos.

necessário

Devolve:

Tipo Descrição
Tuple[Tensor, Tensor, Tensor]

Uma tupla que contém os seguintes tensores: - target_labels (Tensor): Forma (b, hw), contendo os rótulos alvo para pontos de ancoragem positivos. - target_bboxes (Tensor): Forma (b, hw, 4), que contém as caixas delimitadoras alvo para os pontos de ancoragem positivos. - target_scores (Tensor): Forma (b, h*w, num_classes), que contém as pontuações alvo para pontos de ancoragem positivos, onde num_classes é o número de classes de objectos.

Código fonte em ultralytics/utils/tal.py
def get_targets(self, gt_labels, gt_bboxes, target_gt_idx, fg_mask):
    """
    Compute target labels, target bounding boxes, and target scores for the positive anchor points.

    Args:
        gt_labels (Tensor): Ground truth labels of shape (b, max_num_obj, 1), where b is the
                            batch size and max_num_obj is the maximum number of objects.
        gt_bboxes (Tensor): Ground truth bounding boxes of shape (b, max_num_obj, 4).
        target_gt_idx (Tensor): Indices of the assigned ground truth objects for positive
                                anchor points, with shape (b, h*w), where h*w is the total
                                number of anchor points.
        fg_mask (Tensor): A boolean tensor of shape (b, h*w) indicating the positive
                          (foreground) anchor points.

    Returns:
        (Tuple[Tensor, Tensor, Tensor]): A tuple containing the following tensors:
            - target_labels (Tensor): Shape (b, h*w), containing the target labels for
                                      positive anchor points.
            - target_bboxes (Tensor): Shape (b, h*w, 4), containing the target bounding boxes
                                      for positive anchor points.
            - target_scores (Tensor): Shape (b, h*w, num_classes), containing the target scores
                                      for positive anchor points, where num_classes is the number
                                      of object classes.
    """

    # Assigned target labels, (b, 1)
    batch_ind = torch.arange(end=self.bs, dtype=torch.int64, device=gt_labels.device)[..., None]
    target_gt_idx = target_gt_idx + batch_ind * self.n_max_boxes  # (b, h*w)
    target_labels = gt_labels.long().flatten()[target_gt_idx]  # (b, h*w)

    # Assigned target boxes, (b, max_num_obj, 4) -> (b, h*w, 4)
    target_bboxes = gt_bboxes.view(-1, gt_bboxes.shape[-1])[target_gt_idx]

    # Assigned target scores
    target_labels.clamp_(0)

    # 10x faster than F.one_hot()
    target_scores = torch.zeros(
        (target_labels.shape[0], target_labels.shape[1], self.num_classes),
        dtype=torch.int64,
        device=target_labels.device,
    )  # (b, h*w, 80)
    target_scores.scatter_(2, target_labels.unsqueeze(-1), 1)

    fg_scores_mask = fg_mask[:, :, None].repeat(1, 1, self.num_classes)  # (b, h*w, 80)
    target_scores = torch.where(fg_scores_mask > 0, target_scores, 0)

    return target_labels, target_bboxes, target_scores

iou_calculation(gt_bboxes, pd_bboxes)

Cálculo de IoU para caixas delimitadoras horizontais.

Código fonte em ultralytics/utils/tal.py
def iou_calculation(self, gt_bboxes, pd_bboxes):
    """IoU calculation for horizontal bounding boxes."""
    return bbox_iou(gt_bboxes, pd_bboxes, xywh=False, CIoU=True).squeeze(-1).clamp_(0)

select_candidates_in_gts(xy_centers, gt_bboxes, eps=1e-09) staticmethod

Selecciona o centro de ancoragem positivo em gt.

Parâmetros:

Nome Tipo Descrição Predefinição
xy_centers Tensor

forma(h*w, 2)

necessário
gt_bboxes Tensor

forma(b, n_caixas, 4)

necessário

Devolve:

Tipo Descrição
Tensor

forma(b, n_caixas, h*w)

Código fonte em ultralytics/utils/tal.py
@staticmethod
def select_candidates_in_gts(xy_centers, gt_bboxes, eps=1e-9):
    """
    Select the positive anchor center in gt.

    Args:
        xy_centers (Tensor): shape(h*w, 2)
        gt_bboxes (Tensor): shape(b, n_boxes, 4)

    Returns:
        (Tensor): shape(b, n_boxes, h*w)
    """
    n_anchors = xy_centers.shape[0]
    bs, n_boxes, _ = gt_bboxes.shape
    lt, rb = gt_bboxes.view(-1, 1, 4).chunk(2, 2)  # left-top, right-bottom
    bbox_deltas = torch.cat((xy_centers[None] - lt, rb - xy_centers[None]), dim=2).view(bs, n_boxes, n_anchors, -1)
    # return (bbox_deltas.min(3)[0] > eps).to(gt_bboxes.dtype)
    return bbox_deltas.amin(3).gt_(eps)

select_highest_overlaps(mask_pos, overlaps, n_max_boxes) staticmethod

Se uma caixa de ancoragem for atribuída a vários gts, será selecionada a que tiver o IoU mais elevado.

Parâmetros:

Nome Tipo Descrição Predefinição
mask_pos Tensor

forma(b, n_max_boxes, h*w)

necessário
overlaps Tensor

forma(b, n_max_boxes, h*w)

necessário

Devolve:

Nome Tipo Descrição
target_gt_idx Tensor

forma(b, h*w)

fg_mask Tensor

forma(b, h*w)

mask_pos Tensor

forma(b, n_max_boxes, h*w)

Código fonte em ultralytics/utils/tal.py
@staticmethod
def select_highest_overlaps(mask_pos, overlaps, n_max_boxes):
    """
    If an anchor box is assigned to multiple gts, the one with the highest IoU will be selected.

    Args:
        mask_pos (Tensor): shape(b, n_max_boxes, h*w)
        overlaps (Tensor): shape(b, n_max_boxes, h*w)

    Returns:
        target_gt_idx (Tensor): shape(b, h*w)
        fg_mask (Tensor): shape(b, h*w)
        mask_pos (Tensor): shape(b, n_max_boxes, h*w)
    """
    # (b, n_max_boxes, h*w) -> (b, h*w)
    fg_mask = mask_pos.sum(-2)
    if fg_mask.max() > 1:  # one anchor is assigned to multiple gt_bboxes
        mask_multi_gts = (fg_mask.unsqueeze(1) > 1).expand(-1, n_max_boxes, -1)  # (b, n_max_boxes, h*w)
        max_overlaps_idx = overlaps.argmax(1)  # (b, h*w)

        is_max_overlaps = torch.zeros(mask_pos.shape, dtype=mask_pos.dtype, device=mask_pos.device)
        is_max_overlaps.scatter_(1, max_overlaps_idx.unsqueeze(1), 1)

        mask_pos = torch.where(mask_multi_gts, is_max_overlaps, mask_pos).float()  # (b, n_max_boxes, h*w)
        fg_mask = mask_pos.sum(-2)
    # Find each grid serve which gt(index)
    target_gt_idx = mask_pos.argmax(-2)  # (b, h*w)
    return target_gt_idx, fg_mask, mask_pos

select_topk_candidates(metrics, largest=True, topk_mask=None)

Selecciona os candidatos top-k com base nas métricas dadas.

Parâmetros:

Nome Tipo Descrição Predefinição
metrics Tensor

Um tensor de forma (b, max_num_obj, hw), em que b é o tamanho do lote, max_num_obj é o número máximo de objectos e hwrepresenta o representa o número total de pontos de ancoragem.

necessário
largest bool

Se for Verdadeiro, selecciona os valores maiores; caso contrário, selecciona os valores mais pequenos.

True
topk_mask Tensor

Um booleano opcional tensor da forma (b, max_num_obj, topk), em que topk é o número de candidatos principais a considerar. Se não for fornecido, os valores top-k são automaticamente calculados com base nas métricas fornecidas.

None

Devolve:

Tipo Descrição
Tensor

Um tensor com a forma (b, max_num_obj, h*w) que contém os principais candidatos seleccionados.

Código fonte em ultralytics/utils/tal.py
def select_topk_candidates(self, metrics, largest=True, topk_mask=None):
    """
    Select the top-k candidates based on the given metrics.

    Args:
        metrics (Tensor): A tensor of shape (b, max_num_obj, h*w), where b is the batch size,
                          max_num_obj is the maximum number of objects, and h*w represents the
                          total number of anchor points.
        largest (bool): If True, select the largest values; otherwise, select the smallest values.
        topk_mask (Tensor): An optional boolean tensor of shape (b, max_num_obj, topk), where
                            topk is the number of top candidates to consider. If not provided,
                            the top-k values are automatically computed based on the given metrics.

    Returns:
        (Tensor): A tensor of shape (b, max_num_obj, h*w) containing the selected top-k candidates.
    """

    # (b, max_num_obj, topk)
    topk_metrics, topk_idxs = torch.topk(metrics, self.topk, dim=-1, largest=largest)
    if topk_mask is None:
        topk_mask = (topk_metrics.max(-1, keepdim=True)[0] > self.eps).expand_as(topk_idxs)
    # (b, max_num_obj, topk)
    topk_idxs.masked_fill_(~topk_mask, 0)

    # (b, max_num_obj, topk, h*w) -> (b, max_num_obj, h*w)
    count_tensor = torch.zeros(metrics.shape, dtype=torch.int8, device=topk_idxs.device)
    ones = torch.ones_like(topk_idxs[:, :, :1], dtype=torch.int8, device=topk_idxs.device)
    for k in range(self.topk):
        # Expand topk_idxs for each value of k and add 1 at the specified positions
        count_tensor.scatter_add_(-1, topk_idxs[:, :, k : k + 1], ones)
    # count_tensor.scatter_add_(-1, topk_idxs, torch.ones_like(topk_idxs, dtype=torch.int8, device=topk_idxs.device))
    # Filter invalid bboxes
    count_tensor.masked_fill_(count_tensor > 1, 0)

    return count_tensor.to(metrics.dtype)



ultralytics.utils.tal.RotatedTaskAlignedAssigner

Bases: TaskAlignedAssigner

Código fonte em ultralytics/utils/tal.py
class RotatedTaskAlignedAssigner(TaskAlignedAssigner):
    def iou_calculation(self, gt_bboxes, pd_bboxes):
        """IoU calculation for rotated bounding boxes."""
        return probiou(gt_bboxes, pd_bboxes).squeeze(-1).clamp_(0)

    @staticmethod
    def select_candidates_in_gts(xy_centers, gt_bboxes):
        """
        Select the positive anchor center in gt for rotated bounding boxes.

        Args:
            xy_centers (Tensor): shape(h*w, 2)
            gt_bboxes (Tensor): shape(b, n_boxes, 5)

        Returns:
            (Tensor): shape(b, n_boxes, h*w)
        """
        # (b, n_boxes, 5) --> (b, n_boxes, 4, 2)
        corners = xywhr2xyxyxyxy(gt_bboxes)
        # (b, n_boxes, 1, 2)
        a, b, _, d = corners.split(1, dim=-2)
        ab = b - a
        ad = d - a

        # (b, n_boxes, h*w, 2)
        ap = xy_centers - a
        norm_ab = (ab * ab).sum(dim=-1)
        norm_ad = (ad * ad).sum(dim=-1)
        ap_dot_ab = (ap * ab).sum(dim=-1)
        ap_dot_ad = (ap * ad).sum(dim=-1)
        return (ap_dot_ab >= 0) & (ap_dot_ab <= norm_ab) & (ap_dot_ad >= 0) & (ap_dot_ad <= norm_ad)  # is_in_box

iou_calculation(gt_bboxes, pd_bboxes)

Cálculo de IoU para caixas delimitadoras rodadas.

Código fonte em ultralytics/utils/tal.py
def iou_calculation(self, gt_bboxes, pd_bboxes):
    """IoU calculation for rotated bounding boxes."""
    return probiou(gt_bboxes, pd_bboxes).squeeze(-1).clamp_(0)

select_candidates_in_gts(xy_centers, gt_bboxes) staticmethod

Selecciona o centro de ancoragem positivo em gt para caixas delimitadoras rodadas.

Parâmetros:

Nome Tipo Descrição Predefinição
xy_centers Tensor

forma(h*w, 2)

necessário
gt_bboxes Tensor

forma(b, n_caixas, 5)

necessário

Devolve:

Tipo Descrição
Tensor

forma(b, n_caixas, h*w)

Código fonte em ultralytics/utils/tal.py
@staticmethod
def select_candidates_in_gts(xy_centers, gt_bboxes):
    """
    Select the positive anchor center in gt for rotated bounding boxes.

    Args:
        xy_centers (Tensor): shape(h*w, 2)
        gt_bboxes (Tensor): shape(b, n_boxes, 5)

    Returns:
        (Tensor): shape(b, n_boxes, h*w)
    """
    # (b, n_boxes, 5) --> (b, n_boxes, 4, 2)
    corners = xywhr2xyxyxyxy(gt_bboxes)
    # (b, n_boxes, 1, 2)
    a, b, _, d = corners.split(1, dim=-2)
    ab = b - a
    ad = d - a

    # (b, n_boxes, h*w, 2)
    ap = xy_centers - a
    norm_ab = (ab * ab).sum(dim=-1)
    norm_ad = (ad * ad).sum(dim=-1)
    ap_dot_ab = (ap * ab).sum(dim=-1)
    ap_dot_ad = (ap * ad).sum(dim=-1)
    return (ap_dot_ab >= 0) & (ap_dot_ab <= norm_ab) & (ap_dot_ad >= 0) & (ap_dot_ad <= norm_ad)  # is_in_box



ultralytics.utils.tal.make_anchors(feats, strides, grid_cell_offset=0.5)

Gera âncoras a partir de características.

Código fonte em ultralytics/utils/tal.py
def make_anchors(feats, strides, grid_cell_offset=0.5):
    """Generate anchors from features."""
    anchor_points, stride_tensor = [], []
    assert feats is not None
    dtype, device = feats[0].dtype, feats[0].device
    for i, stride in enumerate(strides):
        _, _, h, w = feats[i].shape
        sx = torch.arange(end=w, device=device, dtype=dtype) + grid_cell_offset  # shift x
        sy = torch.arange(end=h, device=device, dtype=dtype) + grid_cell_offset  # shift y
        sy, sx = torch.meshgrid(sy, sx, indexing="ij") if TORCH_1_10 else torch.meshgrid(sy, sx)
        anchor_points.append(torch.stack((sx, sy), -1).view(-1, 2))
        stride_tensor.append(torch.full((h * w, 1), stride, dtype=dtype, device=device))
    return torch.cat(anchor_points), torch.cat(stride_tensor)



ultralytics.utils.tal.dist2bbox(distance, anchor_points, xywh=True, dim=-1)

Transforma a distância (ltrb) em caixa (xywh ou xyxy).

Código fonte em ultralytics/utils/tal.py
def dist2bbox(distance, anchor_points, xywh=True, dim=-1):
    """Transform distance(ltrb) to box(xywh or xyxy)."""
    lt, rb = distance.chunk(2, dim)
    x1y1 = anchor_points - lt
    x2y2 = anchor_points + rb
    if xywh:
        c_xy = (x1y1 + x2y2) / 2
        wh = x2y2 - x1y1
        return torch.cat((c_xy, wh), dim)  # xywh bbox
    return torch.cat((x1y1, x2y2), dim)  # xyxy bbox



ultralytics.utils.tal.bbox2dist(anchor_points, bbox, reg_max)

Transforma bbox(xyxy) em dist(ltrb).

Código fonte em ultralytics/utils/tal.py
def bbox2dist(anchor_points, bbox, reg_max):
    """Transform bbox(xyxy) to dist(ltrb)."""
    x1y1, x2y2 = bbox.chunk(2, -1)
    return torch.cat((anchor_points - x1y1, x2y2 - anchor_points), -1).clamp_(0, reg_max - 0.01)  # dist (lt, rb)



ultralytics.utils.tal.dist2rbox(pred_dist, pred_angle, anchor_points, dim=-1)

Descodifica as coordenadas da caixa delimitadora do objeto previsto a partir dos pontos de ancoragem e da distribuição.

Parâmetros:

Nome Tipo Descrição Predefinição
pred_dist Tensor

Distância rodada prevista, (bs, h*w, 4).

necessário
pred_angle Tensor

Ângulo previsto, (bs, h*w, 1).

necessário
anchor_points Tensor

Pontos de ancoragem, (h*w, 2).

necessário

Devolve: (torch.Tensor): Caixas delimitadoras rotacionadas previstas, (bs, h*w, 4).

Código fonte em ultralytics/utils/tal.py
def dist2rbox(pred_dist, pred_angle, anchor_points, dim=-1):
    """
    Decode predicted object bounding box coordinates from anchor points and distribution.

    Args:
        pred_dist (torch.Tensor): Predicted rotated distance, (bs, h*w, 4).
        pred_angle (torch.Tensor): Predicted angle, (bs, h*w, 1).
        anchor_points (torch.Tensor): Anchor points, (h*w, 2).
    Returns:
        (torch.Tensor): Predicted rotated bounding boxes, (bs, h*w, 4).
    """
    lt, rb = pred_dist.split(2, dim=dim)
    cos, sin = torch.cos(pred_angle), torch.sin(pred_angle)
    # (bs, h*w, 1)
    xf, yf = ((rb - lt) / 2).split(1, dim=dim)
    x, y = xf * cos - yf * sin, xf * sin + yf * cos
    xy = torch.cat([x, y], dim=dim) + anchor_points
    return torch.cat([xy, lt + rb], dim=dim)





Criado em 2023-11-12, Atualizado em 2024-01-05
Autores: glenn-jocher (4), Laughing-q (1)