─░├žeri─če ge├ž

Referans i├žin ultralytics/nn/modules/head.py

Not

Bu dosya https://github.com/ultralytics/ultralytics/blob/main/ ultralytics/nn/modules/head .py adresinde mevcuttur. Bir sorun tespit ederseniz l├╝tfen bir ├çekme ─░ste─či ­čŤá´ŞĆ ile katk─▒da bulunarak d├╝zeltilmesine yard─▒mc─▒ olun. Te┼čekk├╝rler ­čÖĆ!



ultralytics.nn.modules.head.Detect

├ťsler: Module

YOLOv8 Alg─▒lama modelleri i├žin kafay─▒ alg─▒lay─▒n.

Kaynak kodu ultralytics/nn/modules/head.py
class Detect(nn.Module):
    """YOLOv8 Detect head for detection models."""

    dynamic = False  # force grid reconstruction
    export = False  # export mode
    end2end = False  # end2end
    max_det = 300  # max_det
    shape = None
    anchors = torch.empty(0)  # init
    strides = torch.empty(0)  # init

    def __init__(self, nc=80, ch=()):
        """Initializes the YOLOv8 detection layer with specified number of classes and channels."""
        super().__init__()
        self.nc = nc  # number of classes
        self.nl = len(ch)  # number of detection layers
        self.reg_max = 16  # DFL channels (ch[0] // 16 to scale 4/8/12/16/20 for n/s/m/l/x)
        self.no = nc + self.reg_max * 4  # number of outputs per anchor
        self.stride = torch.zeros(self.nl)  # strides computed during build
        c2, c3 = max((16, ch[0] // 4, self.reg_max * 4)), max(ch[0], min(self.nc, 100))  # channels
        self.cv2 = nn.ModuleList(
            nn.Sequential(Conv(x, c2, 3), Conv(c2, c2, 3), nn.Conv2d(c2, 4 * self.reg_max, 1)) for x in ch
        )
        self.cv3 = nn.ModuleList(nn.Sequential(Conv(x, c3, 3), Conv(c3, c3, 3), nn.Conv2d(c3, self.nc, 1)) for x in ch)
        self.dfl = DFL(self.reg_max) if self.reg_max > 1 else nn.Identity()

        if self.end2end:
            self.one2one_cv2 = copy.deepcopy(self.cv2)
            self.one2one_cv3 = copy.deepcopy(self.cv3)

    def forward(self, x):
        """Concatenates and returns predicted bounding boxes and class probabilities."""
        if self.end2end:
            return self.forward_end2end(x)

        for i in range(self.nl):
            x[i] = torch.cat((self.cv2[i](x[i]), self.cv3[i](x[i])), 1)
        if self.training:  # Training path
            return x
        y = self._inference(x)
        return y if self.export else (y, x)

    def forward_end2end(self, x):
        """
        Performs forward pass of the v10Detect module.

        Args:
            x (tensor): Input tensor.

        Returns:
            (dict, tensor): If not in training mode, returns a dictionary containing the outputs of both one2many and one2one detections.
                           If in training mode, returns a dictionary containing the outputs of one2many and one2one detections separately.
        """
        x_detach = [xi.detach() for xi in x]
        one2one = [
            torch.cat((self.one2one_cv2[i](x_detach[i]), self.one2one_cv3[i](x_detach[i])), 1) for i in range(self.nl)
        ]
        for i in range(self.nl):
            x[i] = torch.cat((self.cv2[i](x[i]), self.cv3[i](x[i])), 1)
        if self.training:  # Training path
            return {"one2many": x, "one2one": one2one}

        y = self._inference(one2one)
        y = self.postprocess(y.permute(0, 2, 1), self.max_det, self.nc)
        return y if self.export else (y, {"one2many": x, "one2one": one2one})

    def _inference(self, x):
        """Decode predicted bounding boxes and class probabilities based on multiple-level feature maps."""
        # Inference path
        shape = x[0].shape  # BCHW
        x_cat = torch.cat([xi.view(shape[0], self.no, -1) for xi in x], 2)
        if self.dynamic or self.shape != shape:
            self.anchors, self.strides = (x.transpose(0, 1) for x in make_anchors(x, self.stride, 0.5))
            self.shape = shape

        if self.export and self.format in {"saved_model", "pb", "tflite", "edgetpu", "tfjs"}:  # avoid TF FlexSplitV ops
            box = x_cat[:, : self.reg_max * 4]
            cls = x_cat[:, self.reg_max * 4 :]
        else:
            box, cls = x_cat.split((self.reg_max * 4, self.nc), 1)

        if self.export and self.format in {"tflite", "edgetpu"}:
            # Precompute normalization factor to increase numerical stability
            # See https://github.com/ultralytics/ultralytics/issues/7371
            grid_h = shape[2]
            grid_w = shape[3]
            grid_size = torch.tensor([grid_w, grid_h, grid_w, grid_h], device=box.device).reshape(1, 4, 1)
            norm = self.strides / (self.stride[0] * grid_size)
            dbox = self.decode_bboxes(self.dfl(box) * norm, self.anchors.unsqueeze(0) * norm[:, :2])
        else:
            dbox = self.decode_bboxes(self.dfl(box), self.anchors.unsqueeze(0)) * self.strides

        return torch.cat((dbox, cls.sigmoid()), 1)

    def bias_init(self):
        """Initialize Detect() biases, WARNING: requires stride availability."""
        m = self  # self.model[-1]  # Detect() module
        # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1
        # ncf = math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum())  # nominal class frequency
        for a, b, s in zip(m.cv2, m.cv3, m.stride):  # from
            a[-1].bias.data[:] = 1.0  # box
            b[-1].bias.data[: m.nc] = math.log(5 / m.nc / (640 / s) ** 2)  # cls (.01 objects, 80 classes, 640 img)
        if self.end2end:
            for a, b, s in zip(m.one2one_cv2, m.one2one_cv3, m.stride):  # from
                a[-1].bias.data[:] = 1.0  # box
                b[-1].bias.data[: m.nc] = math.log(5 / m.nc / (640 / s) ** 2)  # cls (.01 objects, 80 classes, 640 img)

    def decode_bboxes(self, bboxes, anchors):
        """Decode bounding boxes."""
        return dist2bbox(bboxes, anchors, xywh=not self.end2end, dim=1)

    @staticmethod
    def postprocess(preds: torch.Tensor, max_det: int, nc: int = 80):
        """
        Post-processes the predictions obtained from a YOLOv10 model.

        Args:
            preds (torch.Tensor): The predictions obtained from the model. It should have a shape of (batch_size, num_boxes, 4 + num_classes).
            max_det (int): The maximum number of detections to keep.
            nc (int, optional): The number of classes. Defaults to 80.

        Returns:
            (torch.Tensor): The post-processed predictions with shape (batch_size, max_det, 6),
                including bounding boxes, scores and cls.
        """
        assert 4 + nc == preds.shape[-1]
        boxes, scores = preds.split([4, nc], dim=-1)
        max_scores = scores.amax(dim=-1)
        max_scores, index = torch.topk(max_scores, min(max_det, max_scores.shape[1]), axis=-1)
        index = index.unsqueeze(-1)
        boxes = torch.gather(boxes, dim=1, index=index.repeat(1, 1, boxes.shape[-1]))
        scores = torch.gather(scores, dim=1, index=index.repeat(1, 1, scores.shape[-1]))

        # NOTE: simplify but result slightly lower mAP
        # scores, labels = scores.max(dim=-1)
        # return torch.cat([boxes, scores.unsqueeze(-1), labels.unsqueeze(-1)], dim=-1)

        scores, index = torch.topk(scores.flatten(1), max_det, axis=-1)
        labels = index % nc
        index = index // nc
        boxes = boxes.gather(dim=1, index=index.unsqueeze(-1).repeat(1, 1, boxes.shape[-1]))

        return torch.cat([boxes, scores.unsqueeze(-1), labels.unsqueeze(-1).to(boxes.dtype)], dim=-1)

__init__(nc=80, ch=())

YOLOv8 alg─▒lama katman─▒n─▒ belirtilen say─▒da s─▒n─▒f ve kanal ile ba┼člat─▒r.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, nc=80, ch=()):
    """Initializes the YOLOv8 detection layer with specified number of classes and channels."""
    super().__init__()
    self.nc = nc  # number of classes
    self.nl = len(ch)  # number of detection layers
    self.reg_max = 16  # DFL channels (ch[0] // 16 to scale 4/8/12/16/20 for n/s/m/l/x)
    self.no = nc + self.reg_max * 4  # number of outputs per anchor
    self.stride = torch.zeros(self.nl)  # strides computed during build
    c2, c3 = max((16, ch[0] // 4, self.reg_max * 4)), max(ch[0], min(self.nc, 100))  # channels
    self.cv2 = nn.ModuleList(
        nn.Sequential(Conv(x, c2, 3), Conv(c2, c2, 3), nn.Conv2d(c2, 4 * self.reg_max, 1)) for x in ch
    )
    self.cv3 = nn.ModuleList(nn.Sequential(Conv(x, c3, 3), Conv(c3, c3, 3), nn.Conv2d(c3, self.nc, 1)) for x in ch)
    self.dfl = DFL(self.reg_max) if self.reg_max > 1 else nn.Identity()

    if self.end2end:
        self.one2one_cv2 = copy.deepcopy(self.cv2)
        self.one2one_cv3 = copy.deepcopy(self.cv3)

bias_init()

Detect() ├Ânyarg─▒lar─▒n─▒ ba┼člat─▒n, UYARI: stride kullan─▒labilirli─či gerektirir.

Kaynak kodu ultralytics/nn/modules/head.py
def bias_init(self):
    """Initialize Detect() biases, WARNING: requires stride availability."""
    m = self  # self.model[-1]  # Detect() module
    # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1
    # ncf = math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum())  # nominal class frequency
    for a, b, s in zip(m.cv2, m.cv3, m.stride):  # from
        a[-1].bias.data[:] = 1.0  # box
        b[-1].bias.data[: m.nc] = math.log(5 / m.nc / (640 / s) ** 2)  # cls (.01 objects, 80 classes, 640 img)
    if self.end2end:
        for a, b, s in zip(m.one2one_cv2, m.one2one_cv3, m.stride):  # from
            a[-1].bias.data[:] = 1.0  # box
            b[-1].bias.data[: m.nc] = math.log(5 / m.nc / (640 / s) ** 2)  # cls (.01 objects, 80 classes, 640 img)

decode_bboxes(bboxes, anchors)

S─▒n─▒rlay─▒c─▒ kutular─▒n kodunu ├ž├Âz├╝n.

Kaynak kodu ultralytics/nn/modules/head.py
def decode_bboxes(self, bboxes, anchors):
    """Decode bounding boxes."""
    return dist2bbox(bboxes, anchors, xywh=not self.end2end, dim=1)

forward(x)

Tahmin edilen s─▒n─▒rlay─▒c─▒ kutular─▒ ve s─▒n─▒f olas─▒l─▒klar─▒n─▒ birle┼čtirir ve d├Ând├╝r├╝r.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x):
    """Concatenates and returns predicted bounding boxes and class probabilities."""
    if self.end2end:
        return self.forward_end2end(x)

    for i in range(self.nl):
        x[i] = torch.cat((self.cv2[i](x[i]), self.cv3[i](x[i])), 1)
    if self.training:  # Training path
        return x
    y = self._inference(x)
    return y if self.export else (y, x)

forward_end2end(x)

Performs forward pass of the v10Detect module.

Parametreler:

─░sim Tip A├ž─▒klama Varsay─▒lan
x tensor

Input tensor.

gerekli

─░ade:

Tip A├ž─▒klama
(dict, tensor)

If not in training mode, returns a dictionary containing the outputs of both one2many and one2one detections. If in training mode, returns a dictionary containing the outputs of one2many and one2one detections separately.

Kaynak kodu ultralytics/nn/modules/head.py
def forward_end2end(self, x):
    """
    Performs forward pass of the v10Detect module.

    Args:
        x (tensor): Input tensor.

    Returns:
        (dict, tensor): If not in training mode, returns a dictionary containing the outputs of both one2many and one2one detections.
                       If in training mode, returns a dictionary containing the outputs of one2many and one2one detections separately.
    """
    x_detach = [xi.detach() for xi in x]
    one2one = [
        torch.cat((self.one2one_cv2[i](x_detach[i]), self.one2one_cv3[i](x_detach[i])), 1) for i in range(self.nl)
    ]
    for i in range(self.nl):
        x[i] = torch.cat((self.cv2[i](x[i]), self.cv3[i](x[i])), 1)
    if self.training:  # Training path
        return {"one2many": x, "one2one": one2one}

    y = self._inference(one2one)
    y = self.postprocess(y.permute(0, 2, 1), self.max_det, self.nc)
    return y if self.export else (y, {"one2many": x, "one2one": one2one})

postprocess(preds, max_det, nc=80) staticmethod

Post-processes the predictions obtained from a YOLOv10 model.

Parametreler:

─░sim Tip A├ž─▒klama Varsay─▒lan
preds Tensor

The predictions obtained from the model. It should have a shape of (batch_size, num_boxes, 4 + num_classes).

gerekli
max_det int

The maximum number of detections to keep.

gerekli
nc int

The number of classes. Defaults to 80.

80

─░ade:

Tip A├ž─▒klama
Tensor

The post-processed predictions with shape (batch_size, max_det, 6), including bounding boxes, scores and cls.

Kaynak kodu ultralytics/nn/modules/head.py
@staticmethod
def postprocess(preds: torch.Tensor, max_det: int, nc: int = 80):
    """
    Post-processes the predictions obtained from a YOLOv10 model.

    Args:
        preds (torch.Tensor): The predictions obtained from the model. It should have a shape of (batch_size, num_boxes, 4 + num_classes).
        max_det (int): The maximum number of detections to keep.
        nc (int, optional): The number of classes. Defaults to 80.

    Returns:
        (torch.Tensor): The post-processed predictions with shape (batch_size, max_det, 6),
            including bounding boxes, scores and cls.
    """
    assert 4 + nc == preds.shape[-1]
    boxes, scores = preds.split([4, nc], dim=-1)
    max_scores = scores.amax(dim=-1)
    max_scores, index = torch.topk(max_scores, min(max_det, max_scores.shape[1]), axis=-1)
    index = index.unsqueeze(-1)
    boxes = torch.gather(boxes, dim=1, index=index.repeat(1, 1, boxes.shape[-1]))
    scores = torch.gather(scores, dim=1, index=index.repeat(1, 1, scores.shape[-1]))

    # NOTE: simplify but result slightly lower mAP
    # scores, labels = scores.max(dim=-1)
    # return torch.cat([boxes, scores.unsqueeze(-1), labels.unsqueeze(-1)], dim=-1)

    scores, index = torch.topk(scores.flatten(1), max_det, axis=-1)
    labels = index % nc
    index = index // nc
    boxes = boxes.gather(dim=1, index=index.unsqueeze(-1).repeat(1, 1, boxes.shape[-1]))

    return torch.cat([boxes, scores.unsqueeze(-1), labels.unsqueeze(-1).to(boxes.dtype)], dim=-1)



ultralytics.nn.modules.head.Segment

├ťsler: Detect

YOLOv8 Segmentasyon modelleri i├žin segment ba┼čl─▒─č─▒.

Kaynak kodu ultralytics/nn/modules/head.py
class Segment(Detect):
    """YOLOv8 Segment head for segmentation models."""

    def __init__(self, nc=80, nm=32, npr=256, ch=()):
        """Initialize the YOLO model attributes such as the number of masks, prototypes, and the convolution layers."""
        super().__init__(nc, ch)
        self.nm = nm  # number of masks
        self.npr = npr  # number of protos
        self.proto = Proto(ch[0], self.npr, self.nm)  # protos

        c4 = max(ch[0] // 4, self.nm)
        self.cv4 = nn.ModuleList(nn.Sequential(Conv(x, c4, 3), Conv(c4, c4, 3), nn.Conv2d(c4, self.nm, 1)) for x in ch)

    def forward(self, x):
        """Return model outputs and mask coefficients if training, otherwise return outputs and mask coefficients."""
        p = self.proto(x[0])  # mask protos
        bs = p.shape[0]  # batch size

        mc = torch.cat([self.cv4[i](x[i]).view(bs, self.nm, -1) for i in range(self.nl)], 2)  # mask coefficients
        x = Detect.forward(self, x)
        if self.training:
            return x, mc, p
        return (torch.cat([x, mc], 1), p) if self.export else (torch.cat([x[0], mc], 1), (x[1], mc, p))

__init__(nc=80, nm=32, npr=256, ch=())

Maske say─▒s─▒, prototipler ve konvol├╝syon katmanlar─▒ gibi YOLO model ├Âzniteliklerini ba┼člat─▒n.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, nc=80, nm=32, npr=256, ch=()):
    """Initialize the YOLO model attributes such as the number of masks, prototypes, and the convolution layers."""
    super().__init__(nc, ch)
    self.nm = nm  # number of masks
    self.npr = npr  # number of protos
    self.proto = Proto(ch[0], self.npr, self.nm)  # protos

    c4 = max(ch[0] // 4, self.nm)
    self.cv4 = nn.ModuleList(nn.Sequential(Conv(x, c4, 3), Conv(c4, c4, 3), nn.Conv2d(c4, self.nm, 1)) for x in ch)

forward(x)

E─čitim varsa model ├ž─▒kt─▒lar─▒n─▒ ve maske katsay─▒lar─▒n─▒ d├Ând├╝r├╝r, aksi takdirde ├ž─▒kt─▒lar─▒ ve maske katsay─▒lar─▒n─▒ d├Ând├╝r├╝r.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x):
    """Return model outputs and mask coefficients if training, otherwise return outputs and mask coefficients."""
    p = self.proto(x[0])  # mask protos
    bs = p.shape[0]  # batch size

    mc = torch.cat([self.cv4[i](x[i]).view(bs, self.nm, -1) for i in range(self.nl)], 2)  # mask coefficients
    x = Detect.forward(self, x)
    if self.training:
        return x, mc, p
    return (torch.cat([x, mc], 1), p) if self.export else (torch.cat([x[0], mc], 1), (x[1], mc, p))



ultralytics.nn.modules.head.OBB

├ťsler: Detect

YOLOv8 Rotasyon modelleri ile alg─▒lama i├žin OBB alg─▒lama ba┼čl─▒─č─▒.

Kaynak kodu ultralytics/nn/modules/head.py
class OBB(Detect):
    """YOLOv8 OBB detection head for detection with rotation models."""

    def __init__(self, nc=80, ne=1, ch=()):
        """Initialize OBB with number of classes `nc` and layer channels `ch`."""
        super().__init__(nc, ch)
        self.ne = ne  # number of extra parameters

        c4 = max(ch[0] // 4, self.ne)
        self.cv4 = nn.ModuleList(nn.Sequential(Conv(x, c4, 3), Conv(c4, c4, 3), nn.Conv2d(c4, self.ne, 1)) for x in ch)

    def forward(self, x):
        """Concatenates and returns predicted bounding boxes and class probabilities."""
        bs = x[0].shape[0]  # batch size
        angle = torch.cat([self.cv4[i](x[i]).view(bs, self.ne, -1) for i in range(self.nl)], 2)  # OBB theta logits
        # NOTE: set `angle` as an attribute so that `decode_bboxes` could use it.
        angle = (angle.sigmoid() - 0.25) * math.pi  # [-pi/4, 3pi/4]
        # angle = angle.sigmoid() * math.pi / 2  # [0, pi/2]
        if not self.training:
            self.angle = angle
        x = Detect.forward(self, x)
        if self.training:
            return x, angle
        return torch.cat([x, angle], 1) if self.export else (torch.cat([x[0], angle], 1), (x[1], angle))

    def decode_bboxes(self, bboxes, anchors):
        """Decode rotated bounding boxes."""
        return dist2rbox(bboxes, self.angle, anchors, dim=1)

__init__(nc=80, ne=1, ch=())

OBB'yi s─▒n─▒f say─▒s─▒ ile ba┼člatma nc ve katman kanallar─▒ ch.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, nc=80, ne=1, ch=()):
    """Initialize OBB with number of classes `nc` and layer channels `ch`."""
    super().__init__(nc, ch)
    self.ne = ne  # number of extra parameters

    c4 = max(ch[0] // 4, self.ne)
    self.cv4 = nn.ModuleList(nn.Sequential(Conv(x, c4, 3), Conv(c4, c4, 3), nn.Conv2d(c4, self.ne, 1)) for x in ch)

decode_bboxes(bboxes, anchors)

D├Ând├╝r├╝lm├╝┼č s─▒n─▒rlay─▒c─▒ kutular─▒n kodunu ├ž├Âz├╝n.

Kaynak kodu ultralytics/nn/modules/head.py
def decode_bboxes(self, bboxes, anchors):
    """Decode rotated bounding boxes."""
    return dist2rbox(bboxes, self.angle, anchors, dim=1)

forward(x)

Tahmin edilen s─▒n─▒rlay─▒c─▒ kutular─▒ ve s─▒n─▒f olas─▒l─▒klar─▒n─▒ birle┼čtirir ve d├Ând├╝r├╝r.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x):
    """Concatenates and returns predicted bounding boxes and class probabilities."""
    bs = x[0].shape[0]  # batch size
    angle = torch.cat([self.cv4[i](x[i]).view(bs, self.ne, -1) for i in range(self.nl)], 2)  # OBB theta logits
    # NOTE: set `angle` as an attribute so that `decode_bboxes` could use it.
    angle = (angle.sigmoid() - 0.25) * math.pi  # [-pi/4, 3pi/4]
    # angle = angle.sigmoid() * math.pi / 2  # [0, pi/2]
    if not self.training:
        self.angle = angle
    x = Detect.forward(self, x)
    if self.training:
        return x, angle
    return torch.cat([x, angle], 1) if self.export else (torch.cat([x[0], angle], 1), (x[1], angle))



ultralytics.nn.modules.head.Pose

├ťsler: Detect

YOLOv8 Anahtar nokta modelleri i├žin poz kafas─▒.

Kaynak kodu ultralytics/nn/modules/head.py
class Pose(Detect):
    """YOLOv8 Pose head for keypoints models."""

    def __init__(self, nc=80, kpt_shape=(17, 3), ch=()):
        """Initialize YOLO network with default parameters and Convolutional Layers."""
        super().__init__(nc, ch)
        self.kpt_shape = kpt_shape  # number of keypoints, number of dims (2 for x,y or 3 for x,y,visible)
        self.nk = kpt_shape[0] * kpt_shape[1]  # number of keypoints total

        c4 = max(ch[0] // 4, self.nk)
        self.cv4 = nn.ModuleList(nn.Sequential(Conv(x, c4, 3), Conv(c4, c4, 3), nn.Conv2d(c4, self.nk, 1)) for x in ch)

    def forward(self, x):
        """Perform forward pass through YOLO model and return predictions."""
        bs = x[0].shape[0]  # batch size
        kpt = torch.cat([self.cv4[i](x[i]).view(bs, self.nk, -1) for i in range(self.nl)], -1)  # (bs, 17*3, h*w)
        x = Detect.forward(self, x)
        if self.training:
            return x, kpt
        pred_kpt = self.kpts_decode(bs, kpt)
        return torch.cat([x, pred_kpt], 1) if self.export else (torch.cat([x[0], pred_kpt], 1), (x[1], kpt))

    def kpts_decode(self, bs, kpts):
        """Decodes keypoints."""
        ndim = self.kpt_shape[1]
        if self.export:  # required for TFLite export to avoid 'PLACEHOLDER_FOR_GREATER_OP_CODES' bug
            y = kpts.view(bs, *self.kpt_shape, -1)
            a = (y[:, :, :2] * 2.0 + (self.anchors - 0.5)) * self.strides
            if ndim == 3:
                a = torch.cat((a, y[:, :, 2:3].sigmoid()), 2)
            return a.view(bs, self.nk, -1)
        else:
            y = kpts.clone()
            if ndim == 3:
                y[:, 2::3] = y[:, 2::3].sigmoid()  # sigmoid (WARNING: inplace .sigmoid_() Apple MPS bug)
            y[:, 0::ndim] = (y[:, 0::ndim] * 2.0 + (self.anchors[0] - 0.5)) * self.strides
            y[:, 1::ndim] = (y[:, 1::ndim] * 2.0 + (self.anchors[1] - 0.5)) * self.strides
            return y

__init__(nc=80, kpt_shape=(17, 3), ch=())

YOLO a─č─▒n─▒ varsay─▒lan parametreler ve Evri┼čimsel Katmanlar ile ba┼člat─▒n.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, nc=80, kpt_shape=(17, 3), ch=()):
    """Initialize YOLO network with default parameters and Convolutional Layers."""
    super().__init__(nc, ch)
    self.kpt_shape = kpt_shape  # number of keypoints, number of dims (2 for x,y or 3 for x,y,visible)
    self.nk = kpt_shape[0] * kpt_shape[1]  # number of keypoints total

    c4 = max(ch[0] // 4, self.nk)
    self.cv4 = nn.ModuleList(nn.Sequential(Conv(x, c4, 3), Conv(c4, c4, 3), nn.Conv2d(c4, self.nk, 1)) for x in ch)

forward(x)

YOLO modeli ├╝zerinden ileri ge├ži┼č ger├žekle┼čtirin ve tahminleri d├Ând├╝r├╝n.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x):
    """Perform forward pass through YOLO model and return predictions."""
    bs = x[0].shape[0]  # batch size
    kpt = torch.cat([self.cv4[i](x[i]).view(bs, self.nk, -1) for i in range(self.nl)], -1)  # (bs, 17*3, h*w)
    x = Detect.forward(self, x)
    if self.training:
        return x, kpt
    pred_kpt = self.kpts_decode(bs, kpt)
    return torch.cat([x, pred_kpt], 1) if self.export else (torch.cat([x[0], pred_kpt], 1), (x[1], kpt))

kpts_decode(bs, kpts)

Anahtar noktalar─▒n kodunu ├ž├Âzer.

Kaynak kodu ultralytics/nn/modules/head.py
def kpts_decode(self, bs, kpts):
    """Decodes keypoints."""
    ndim = self.kpt_shape[1]
    if self.export:  # required for TFLite export to avoid 'PLACEHOLDER_FOR_GREATER_OP_CODES' bug
        y = kpts.view(bs, *self.kpt_shape, -1)
        a = (y[:, :, :2] * 2.0 + (self.anchors - 0.5)) * self.strides
        if ndim == 3:
            a = torch.cat((a, y[:, :, 2:3].sigmoid()), 2)
        return a.view(bs, self.nk, -1)
    else:
        y = kpts.clone()
        if ndim == 3:
            y[:, 2::3] = y[:, 2::3].sigmoid()  # sigmoid (WARNING: inplace .sigmoid_() Apple MPS bug)
        y[:, 0::ndim] = (y[:, 0::ndim] * 2.0 + (self.anchors[0] - 0.5)) * self.strides
        y[:, 1::ndim] = (y[:, 1::ndim] * 2.0 + (self.anchors[1] - 0.5)) * self.strides
        return y



ultralytics.nn.modules.head.Classify

├ťsler: Module

YOLOv8 s─▒n─▒fland─▒rma ba┼čl─▒─č─▒, yani x(b,c1,20,20) ila x(b,c2).

Kaynak kodu ultralytics/nn/modules/head.py
class Classify(nn.Module):
    """YOLOv8 classification head, i.e. x(b,c1,20,20) to x(b,c2)."""

    def __init__(self, c1, c2, k=1, s=1, p=None, g=1):
        """Initializes YOLOv8 classification head with specified input and output channels, kernel size, stride,
        padding, and groups.
        """
        super().__init__()
        c_ = 1280  # efficientnet_b0 size
        self.conv = Conv(c1, c_, k, s, p, g)
        self.pool = nn.AdaptiveAvgPool2d(1)  # to x(b,c_,1,1)
        self.drop = nn.Dropout(p=0.0, inplace=True)
        self.linear = nn.Linear(c_, c2)  # to x(b,c2)

    def forward(self, x):
        """Performs a forward pass of the YOLO model on input image data."""
        if isinstance(x, list):
            x = torch.cat(x, 1)
        x = self.linear(self.drop(self.pool(self.conv(x)).flatten(1)))
        return x if self.training else x.softmax(1)

__init__(c1, c2, k=1, s=1, p=None, g=1)

YOLOv8 s─▒n─▒fland─▒rma kafas─▒n─▒ belirtilen giri┼č ve ├ž─▒k─▒┼č kanallar─▒, ├žekirdek boyutu ve stride ile ba┼člat─▒r, dolgu ve gruplar.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, c1, c2, k=1, s=1, p=None, g=1):
    """Initializes YOLOv8 classification head with specified input and output channels, kernel size, stride,
    padding, and groups.
    """
    super().__init__()
    c_ = 1280  # efficientnet_b0 size
    self.conv = Conv(c1, c_, k, s, p, g)
    self.pool = nn.AdaptiveAvgPool2d(1)  # to x(b,c_,1,1)
    self.drop = nn.Dropout(p=0.0, inplace=True)
    self.linear = nn.Linear(c_, c2)  # to x(b,c2)

forward(x)

Giri┼č g├Âr├╝nt├╝ verileri ├╝zerinde YOLO modelinin ileri ge├ži┼čini ger├žekle┼čtirir.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x):
    """Performs a forward pass of the YOLO model on input image data."""
    if isinstance(x, list):
        x = torch.cat(x, 1)
    x = self.linear(self.drop(self.pool(self.conv(x)).flatten(1)))
    return x if self.training else x.softmax(1)



ultralytics.nn.modules.head.WorldDetect

├ťsler: Detect

Kaynak kodu ultralytics/nn/modules/head.py
class WorldDetect(Detect):
    def __init__(self, nc=80, embed=512, with_bn=False, ch=()):
        """Initialize YOLOv8 detection layer with nc classes and layer channels ch."""
        super().__init__(nc, ch)
        c3 = max(ch[0], min(self.nc, 100))
        self.cv3 = nn.ModuleList(nn.Sequential(Conv(x, c3, 3), Conv(c3, c3, 3), nn.Conv2d(c3, embed, 1)) for x in ch)
        self.cv4 = nn.ModuleList(BNContrastiveHead(embed) if with_bn else ContrastiveHead() for _ in ch)

    def forward(self, x, text):
        """Concatenates and returns predicted bounding boxes and class probabilities."""
        for i in range(self.nl):
            x[i] = torch.cat((self.cv2[i](x[i]), self.cv4[i](self.cv3[i](x[i]), text)), 1)
        if self.training:
            return x

        # Inference path
        shape = x[0].shape  # BCHW
        x_cat = torch.cat([xi.view(shape[0], self.nc + self.reg_max * 4, -1) for xi in x], 2)
        if self.dynamic or self.shape != shape:
            self.anchors, self.strides = (x.transpose(0, 1) for x in make_anchors(x, self.stride, 0.5))
            self.shape = shape

        if self.export and self.format in {"saved_model", "pb", "tflite", "edgetpu", "tfjs"}:  # avoid TF FlexSplitV ops
            box = x_cat[:, : self.reg_max * 4]
            cls = x_cat[:, self.reg_max * 4 :]
        else:
            box, cls = x_cat.split((self.reg_max * 4, self.nc), 1)

        if self.export and self.format in {"tflite", "edgetpu"}:
            # Precompute normalization factor to increase numerical stability
            # See https://github.com/ultralytics/ultralytics/issues/7371
            grid_h = shape[2]
            grid_w = shape[3]
            grid_size = torch.tensor([grid_w, grid_h, grid_w, grid_h], device=box.device).reshape(1, 4, 1)
            norm = self.strides / (self.stride[0] * grid_size)
            dbox = self.decode_bboxes(self.dfl(box) * norm, self.anchors.unsqueeze(0) * norm[:, :2])
        else:
            dbox = self.decode_bboxes(self.dfl(box), self.anchors.unsqueeze(0)) * self.strides

        y = torch.cat((dbox, cls.sigmoid()), 1)
        return y if self.export else (y, x)

    def bias_init(self):
        """Initialize Detect() biases, WARNING: requires stride availability."""
        m = self  # self.model[-1]  # Detect() module
        # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1
        # ncf = math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum())  # nominal class frequency
        for a, b, s in zip(m.cv2, m.cv3, m.stride):  # from
            a[-1].bias.data[:] = 1.0  # box

__init__(nc=80, embed=512, with_bn=False, ch=())

YOLOv8 alg─▒lama katman─▒n─▒ nc s─▒n─▒flar─▒ ve ch katman kanallar─▒ ile ba┼člat─▒n.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, nc=80, embed=512, with_bn=False, ch=()):
    """Initialize YOLOv8 detection layer with nc classes and layer channels ch."""
    super().__init__(nc, ch)
    c3 = max(ch[0], min(self.nc, 100))
    self.cv3 = nn.ModuleList(nn.Sequential(Conv(x, c3, 3), Conv(c3, c3, 3), nn.Conv2d(c3, embed, 1)) for x in ch)
    self.cv4 = nn.ModuleList(BNContrastiveHead(embed) if with_bn else ContrastiveHead() for _ in ch)

bias_init()

Detect() ├Ânyarg─▒lar─▒n─▒ ba┼člat─▒n, UYARI: stride kullan─▒labilirli─či gerektirir.

Kaynak kodu ultralytics/nn/modules/head.py
def bias_init(self):
    """Initialize Detect() biases, WARNING: requires stride availability."""
    m = self  # self.model[-1]  # Detect() module
    # cf = torch.bincount(torch.tensor(np.concatenate(dataset.labels, 0)[:, 0]).long(), minlength=nc) + 1
    # ncf = math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum())  # nominal class frequency
    for a, b, s in zip(m.cv2, m.cv3, m.stride):  # from
        a[-1].bias.data[:] = 1.0  # box

forward(x, text)

Tahmin edilen s─▒n─▒rlay─▒c─▒ kutular─▒ ve s─▒n─▒f olas─▒l─▒klar─▒n─▒ birle┼čtirir ve d├Ând├╝r├╝r.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x, text):
    """Concatenates and returns predicted bounding boxes and class probabilities."""
    for i in range(self.nl):
        x[i] = torch.cat((self.cv2[i](x[i]), self.cv4[i](self.cv3[i](x[i]), text)), 1)
    if self.training:
        return x

    # Inference path
    shape = x[0].shape  # BCHW
    x_cat = torch.cat([xi.view(shape[0], self.nc + self.reg_max * 4, -1) for xi in x], 2)
    if self.dynamic or self.shape != shape:
        self.anchors, self.strides = (x.transpose(0, 1) for x in make_anchors(x, self.stride, 0.5))
        self.shape = shape

    if self.export and self.format in {"saved_model", "pb", "tflite", "edgetpu", "tfjs"}:  # avoid TF FlexSplitV ops
        box = x_cat[:, : self.reg_max * 4]
        cls = x_cat[:, self.reg_max * 4 :]
    else:
        box, cls = x_cat.split((self.reg_max * 4, self.nc), 1)

    if self.export and self.format in {"tflite", "edgetpu"}:
        # Precompute normalization factor to increase numerical stability
        # See https://github.com/ultralytics/ultralytics/issues/7371
        grid_h = shape[2]
        grid_w = shape[3]
        grid_size = torch.tensor([grid_w, grid_h, grid_w, grid_h], device=box.device).reshape(1, 4, 1)
        norm = self.strides / (self.stride[0] * grid_size)
        dbox = self.decode_bboxes(self.dfl(box) * norm, self.anchors.unsqueeze(0) * norm[:, :2])
    else:
        dbox = self.decode_bboxes(self.dfl(box), self.anchors.unsqueeze(0)) * self.strides

    y = torch.cat((dbox, cls.sigmoid()), 1)
    return y if self.export else (y, x)



ultralytics.nn.modules.head.RTDETRDecoder

├ťsler: Module

Nesne alg─▒lama i├žin Ger├žek Zamanl─▒ Deforme Edilebilir Transformat├Âr Kod ├ç├Âz├╝c├╝ (RTDETRDecoder) mod├╝l├╝.

Bu kod ├ž├Âz├╝c├╝ mod├╝l├╝, s─▒n─▒rlay─▒c─▒ kutular─▒ tahmin etmek i├žin deforme edilebilir konvol├╝syonlarla birlikte Transformer mimarisini kullan─▒r ve bir g├Âr├╝nt├╝deki nesneler i├žin s─▒n─▒f etiketleri. Birden fazla katmandan gelen ├Âzellikleri entegre eder ve bir dizi Nihai tahminlerin ├ž─▒kt─▒s─▒n─▒ almak i├žin d├Ân├╝┼čt├╝r├╝c├╝ kod ├ž├Âz├╝c├╝ katmanlar─▒.

Kaynak kodu ultralytics/nn/modules/head.py
class RTDETRDecoder(nn.Module):
    """
    Real-Time Deformable Transformer Decoder (RTDETRDecoder) module for object detection.

    This decoder module utilizes Transformer architecture along with deformable convolutions to predict bounding boxes
    and class labels for objects in an image. It integrates features from multiple layers and runs through a series of
    Transformer decoder layers to output the final predictions.
    """

    export = False  # export mode

    def __init__(
        self,
        nc=80,
        ch=(512, 1024, 2048),
        hd=256,  # hidden dim
        nq=300,  # num queries
        ndp=4,  # num decoder points
        nh=8,  # num head
        ndl=6,  # num decoder layers
        d_ffn=1024,  # dim of feedforward
        dropout=0.0,
        act=nn.ReLU(),
        eval_idx=-1,
        # Training args
        nd=100,  # num denoising
        label_noise_ratio=0.5,
        box_noise_scale=1.0,
        learnt_init_query=False,
    ):
        """
        Initializes the RTDETRDecoder module with the given parameters.

        Args:
            nc (int): Number of classes. Default is 80.
            ch (tuple): Channels in the backbone feature maps. Default is (512, 1024, 2048).
            hd (int): Dimension of hidden layers. Default is 256.
            nq (int): Number of query points. Default is 300.
            ndp (int): Number of decoder points. Default is 4.
            nh (int): Number of heads in multi-head attention. Default is 8.
            ndl (int): Number of decoder layers. Default is 6.
            d_ffn (int): Dimension of the feed-forward networks. Default is 1024.
            dropout (float): Dropout rate. Default is 0.
            act (nn.Module): Activation function. Default is nn.ReLU.
            eval_idx (int): Evaluation index. Default is -1.
            nd (int): Number of denoising. Default is 100.
            label_noise_ratio (float): Label noise ratio. Default is 0.5.
            box_noise_scale (float): Box noise scale. Default is 1.0.
            learnt_init_query (bool): Whether to learn initial query embeddings. Default is False.
        """
        super().__init__()
        self.hidden_dim = hd
        self.nhead = nh
        self.nl = len(ch)  # num level
        self.nc = nc
        self.num_queries = nq
        self.num_decoder_layers = ndl

        # Backbone feature projection
        self.input_proj = nn.ModuleList(nn.Sequential(nn.Conv2d(x, hd, 1, bias=False), nn.BatchNorm2d(hd)) for x in ch)
        # NOTE: simplified version but it's not consistent with .pt weights.
        # self.input_proj = nn.ModuleList(Conv(x, hd, act=False) for x in ch)

        # Transformer module
        decoder_layer = DeformableTransformerDecoderLayer(hd, nh, d_ffn, dropout, act, self.nl, ndp)
        self.decoder = DeformableTransformerDecoder(hd, decoder_layer, ndl, eval_idx)

        # Denoising part
        self.denoising_class_embed = nn.Embedding(nc, hd)
        self.num_denoising = nd
        self.label_noise_ratio = label_noise_ratio
        self.box_noise_scale = box_noise_scale

        # Decoder embedding
        self.learnt_init_query = learnt_init_query
        if learnt_init_query:
            self.tgt_embed = nn.Embedding(nq, hd)
        self.query_pos_head = MLP(4, 2 * hd, hd, num_layers=2)

        # Encoder head
        self.enc_output = nn.Sequential(nn.Linear(hd, hd), nn.LayerNorm(hd))
        self.enc_score_head = nn.Linear(hd, nc)
        self.enc_bbox_head = MLP(hd, hd, 4, num_layers=3)

        # Decoder head
        self.dec_score_head = nn.ModuleList([nn.Linear(hd, nc) for _ in range(ndl)])
        self.dec_bbox_head = nn.ModuleList([MLP(hd, hd, 4, num_layers=3) for _ in range(ndl)])

        self._reset_parameters()

    def forward(self, x, batch=None):
        """Runs the forward pass of the module, returning bounding box and classification scores for the input."""
        from ultralytics.models.utils.ops import get_cdn_group

        # Input projection and embedding
        feats, shapes = self._get_encoder_input(x)

        # Prepare denoising training
        dn_embed, dn_bbox, attn_mask, dn_meta = get_cdn_group(
            batch,
            self.nc,
            self.num_queries,
            self.denoising_class_embed.weight,
            self.num_denoising,
            self.label_noise_ratio,
            self.box_noise_scale,
            self.training,
        )

        embed, refer_bbox, enc_bboxes, enc_scores = self._get_decoder_input(feats, shapes, dn_embed, dn_bbox)

        # Decoder
        dec_bboxes, dec_scores = self.decoder(
            embed,
            refer_bbox,
            feats,
            shapes,
            self.dec_bbox_head,
            self.dec_score_head,
            self.query_pos_head,
            attn_mask=attn_mask,
        )
        x = dec_bboxes, dec_scores, enc_bboxes, enc_scores, dn_meta
        if self.training:
            return x
        # (bs, 300, 4+nc)
        y = torch.cat((dec_bboxes.squeeze(0), dec_scores.squeeze(0).sigmoid()), -1)
        return y if self.export else (y, x)

    def _generate_anchors(self, shapes, grid_size=0.05, dtype=torch.float32, device="cpu", eps=1e-2):
        """Generates anchor bounding boxes for given shapes with specific grid size and validates them."""
        anchors = []
        for i, (h, w) in enumerate(shapes):
            sy = torch.arange(end=h, dtype=dtype, device=device)
            sx = torch.arange(end=w, dtype=dtype, device=device)
            grid_y, grid_x = torch.meshgrid(sy, sx, indexing="ij") if TORCH_1_10 else torch.meshgrid(sy, sx)
            grid_xy = torch.stack([grid_x, grid_y], -1)  # (h, w, 2)

            valid_WH = torch.tensor([w, h], dtype=dtype, device=device)
            grid_xy = (grid_xy.unsqueeze(0) + 0.5) / valid_WH  # (1, h, w, 2)
            wh = torch.ones_like(grid_xy, dtype=dtype, device=device) * grid_size * (2.0**i)
            anchors.append(torch.cat([grid_xy, wh], -1).view(-1, h * w, 4))  # (1, h*w, 4)

        anchors = torch.cat(anchors, 1)  # (1, h*w*nl, 4)
        valid_mask = ((anchors > eps) & (anchors < 1 - eps)).all(-1, keepdim=True)  # 1, h*w*nl, 1
        anchors = torch.log(anchors / (1 - anchors))
        anchors = anchors.masked_fill(~valid_mask, float("inf"))
        return anchors, valid_mask

    def _get_encoder_input(self, x):
        """Processes and returns encoder inputs by getting projection features from input and concatenating them."""
        # Get projection features
        x = [self.input_proj[i](feat) for i, feat in enumerate(x)]
        # Get encoder inputs
        feats = []
        shapes = []
        for feat in x:
            h, w = feat.shape[2:]
            # [b, c, h, w] -> [b, h*w, c]
            feats.append(feat.flatten(2).permute(0, 2, 1))
            # [nl, 2]
            shapes.append([h, w])

        # [b, h*w, c]
        feats = torch.cat(feats, 1)
        return feats, shapes

    def _get_decoder_input(self, feats, shapes, dn_embed=None, dn_bbox=None):
        """Generates and prepares the input required for the decoder from the provided features and shapes."""
        bs = feats.shape[0]
        # Prepare input for decoder
        anchors, valid_mask = self._generate_anchors(shapes, dtype=feats.dtype, device=feats.device)
        features = self.enc_output(valid_mask * feats)  # bs, h*w, 256

        enc_outputs_scores = self.enc_score_head(features)  # (bs, h*w, nc)

        # Query selection
        # (bs, num_queries)
        topk_ind = torch.topk(enc_outputs_scores.max(-1).values, self.num_queries, dim=1).indices.view(-1)
        # (bs, num_queries)
        batch_ind = torch.arange(end=bs, dtype=topk_ind.dtype).unsqueeze(-1).repeat(1, self.num_queries).view(-1)

        # (bs, num_queries, 256)
        top_k_features = features[batch_ind, topk_ind].view(bs, self.num_queries, -1)
        # (bs, num_queries, 4)
        top_k_anchors = anchors[:, topk_ind].view(bs, self.num_queries, -1)

        # Dynamic anchors + static content
        refer_bbox = self.enc_bbox_head(top_k_features) + top_k_anchors

        enc_bboxes = refer_bbox.sigmoid()
        if dn_bbox is not None:
            refer_bbox = torch.cat([dn_bbox, refer_bbox], 1)
        enc_scores = enc_outputs_scores[batch_ind, topk_ind].view(bs, self.num_queries, -1)

        embeddings = self.tgt_embed.weight.unsqueeze(0).repeat(bs, 1, 1) if self.learnt_init_query else top_k_features
        if self.training:
            refer_bbox = refer_bbox.detach()
            if not self.learnt_init_query:
                embeddings = embeddings.detach()
        if dn_embed is not None:
            embeddings = torch.cat([dn_embed, embeddings], 1)

        return embeddings, refer_bbox, enc_bboxes, enc_scores

    # TODO
    def _reset_parameters(self):
        """Initializes or resets the parameters of the model's various components with predefined weights and biases."""
        # Class and bbox head init
        bias_cls = bias_init_with_prob(0.01) / 80 * self.nc
        # NOTE: the weight initialization in `linear_init` would cause NaN when training with custom datasets.
        # linear_init(self.enc_score_head)
        constant_(self.enc_score_head.bias, bias_cls)
        constant_(self.enc_bbox_head.layers[-1].weight, 0.0)
        constant_(self.enc_bbox_head.layers[-1].bias, 0.0)
        for cls_, reg_ in zip(self.dec_score_head, self.dec_bbox_head):
            # linear_init(cls_)
            constant_(cls_.bias, bias_cls)
            constant_(reg_.layers[-1].weight, 0.0)
            constant_(reg_.layers[-1].bias, 0.0)

        linear_init(self.enc_output[0])
        xavier_uniform_(self.enc_output[0].weight)
        if self.learnt_init_query:
            xavier_uniform_(self.tgt_embed.weight)
        xavier_uniform_(self.query_pos_head.layers[0].weight)
        xavier_uniform_(self.query_pos_head.layers[1].weight)
        for layer in self.input_proj:
            xavier_uniform_(layer[0].weight)

__init__(nc=80, ch=(512, 1024, 2048), hd=256, nq=300, ndp=4, nh=8, ndl=6, d_ffn=1024, dropout=0.0, act=nn.ReLU(), eval_idx=-1, nd=100, label_noise_ratio=0.5, box_noise_scale=1.0, learnt_init_query=False)

RTDETRDecoder mod├╝l├╝n├╝ verilen parametrelerle ba┼člat─▒r.

Parametreler:

─░sim Tip A├ž─▒klama Varsay─▒lan
nc int

S─▒n─▒f say─▒s─▒. Varsay─▒lan de─čer 80'dir.

80
ch tuple

Omurga ├Âzellik haritalar─▒ndaki kanallar. Varsay─▒lan de─čer (512, 1024, 2048)'dir.

(512, 1024, 2048)
hd int

Gizli katmanlar─▒n boyutu. Varsay─▒lan de─čer 256'd─▒r.

256
nq int

Sorgu noktalar─▒n─▒n say─▒s─▒. Varsay─▒lan de─čer 300'd├╝r.

300
ndp int

Kod ├ž├Âz├╝c├╝ nokta say─▒s─▒. Varsay─▒lan de─čer 4't├╝r.

4
nh int

├çoklu kafa dikkatindeki kafa say─▒s─▒. Varsay─▒lan de─čer 8'dir.

8
ndl int

Kod ├ž├Âz├╝c├╝ katmanlar─▒n─▒n say─▒s─▒. Varsay─▒lan de─čer 6'd─▒r.

6
d_ffn int

─░leri beslemeli a─člar─▒n boyutu. Varsay─▒lan de─čer 1024't├╝r.

1024
dropout float

B─▒rakma oran─▒. Varsay─▒lan de─čer 0'd─▒r.

0.0
act Module

Etkinle┼čtirme i┼člevi. Varsay─▒lan de─čer nn.ReLU'dur.

ReLU()
eval_idx int

De─čerlendirme indeksi. Varsay─▒lan de─čer -1'dir.

-1
nd int

Denoising say─▒s─▒. Varsay─▒lan de─čer 100'd├╝r.

100
label_noise_ratio float

Etiket g├╝r├╝lt├╝ oran─▒. Varsay─▒lan de─čer 0,5'tir.

0.5
box_noise_scale float

Kutu g├╝r├╝lt├╝ ├Âl├že─či. Varsay─▒lan de─čer 1,0'd─▒r.

1.0
learnt_init_query bool

─░lk sorgu kat─▒┼čt─▒rmalar─▒n─▒n ├Â─črenilip ├Â─črenilmeyece─či. Varsay─▒lan de─čer False'dir.

False
Kaynak kodu ultralytics/nn/modules/head.py
def __init__(
    self,
    nc=80,
    ch=(512, 1024, 2048),
    hd=256,  # hidden dim
    nq=300,  # num queries
    ndp=4,  # num decoder points
    nh=8,  # num head
    ndl=6,  # num decoder layers
    d_ffn=1024,  # dim of feedforward
    dropout=0.0,
    act=nn.ReLU(),
    eval_idx=-1,
    # Training args
    nd=100,  # num denoising
    label_noise_ratio=0.5,
    box_noise_scale=1.0,
    learnt_init_query=False,
):
    """
    Initializes the RTDETRDecoder module with the given parameters.

    Args:
        nc (int): Number of classes. Default is 80.
        ch (tuple): Channels in the backbone feature maps. Default is (512, 1024, 2048).
        hd (int): Dimension of hidden layers. Default is 256.
        nq (int): Number of query points. Default is 300.
        ndp (int): Number of decoder points. Default is 4.
        nh (int): Number of heads in multi-head attention. Default is 8.
        ndl (int): Number of decoder layers. Default is 6.
        d_ffn (int): Dimension of the feed-forward networks. Default is 1024.
        dropout (float): Dropout rate. Default is 0.
        act (nn.Module): Activation function. Default is nn.ReLU.
        eval_idx (int): Evaluation index. Default is -1.
        nd (int): Number of denoising. Default is 100.
        label_noise_ratio (float): Label noise ratio. Default is 0.5.
        box_noise_scale (float): Box noise scale. Default is 1.0.
        learnt_init_query (bool): Whether to learn initial query embeddings. Default is False.
    """
    super().__init__()
    self.hidden_dim = hd
    self.nhead = nh
    self.nl = len(ch)  # num level
    self.nc = nc
    self.num_queries = nq
    self.num_decoder_layers = ndl

    # Backbone feature projection
    self.input_proj = nn.ModuleList(nn.Sequential(nn.Conv2d(x, hd, 1, bias=False), nn.BatchNorm2d(hd)) for x in ch)
    # NOTE: simplified version but it's not consistent with .pt weights.
    # self.input_proj = nn.ModuleList(Conv(x, hd, act=False) for x in ch)

    # Transformer module
    decoder_layer = DeformableTransformerDecoderLayer(hd, nh, d_ffn, dropout, act, self.nl, ndp)
    self.decoder = DeformableTransformerDecoder(hd, decoder_layer, ndl, eval_idx)

    # Denoising part
    self.denoising_class_embed = nn.Embedding(nc, hd)
    self.num_denoising = nd
    self.label_noise_ratio = label_noise_ratio
    self.box_noise_scale = box_noise_scale

    # Decoder embedding
    self.learnt_init_query = learnt_init_query
    if learnt_init_query:
        self.tgt_embed = nn.Embedding(nq, hd)
    self.query_pos_head = MLP(4, 2 * hd, hd, num_layers=2)

    # Encoder head
    self.enc_output = nn.Sequential(nn.Linear(hd, hd), nn.LayerNorm(hd))
    self.enc_score_head = nn.Linear(hd, nc)
    self.enc_bbox_head = MLP(hd, hd, 4, num_layers=3)

    # Decoder head
    self.dec_score_head = nn.ModuleList([nn.Linear(hd, nc) for _ in range(ndl)])
    self.dec_bbox_head = nn.ModuleList([MLP(hd, hd, 4, num_layers=3) for _ in range(ndl)])

    self._reset_parameters()

forward(x, batch=None)

Mod├╝l├╝n ileri ge├ži┼čini ├žal─▒┼čt─▒rarak girdi i├žin s─▒n─▒rlay─▒c─▒ kutu ve s─▒n─▒fland─▒rma puanlar─▒n─▒ d├Ând├╝r├╝r.

Kaynak kodu ultralytics/nn/modules/head.py
def forward(self, x, batch=None):
    """Runs the forward pass of the module, returning bounding box and classification scores for the input."""
    from ultralytics.models.utils.ops import get_cdn_group

    # Input projection and embedding
    feats, shapes = self._get_encoder_input(x)

    # Prepare denoising training
    dn_embed, dn_bbox, attn_mask, dn_meta = get_cdn_group(
        batch,
        self.nc,
        self.num_queries,
        self.denoising_class_embed.weight,
        self.num_denoising,
        self.label_noise_ratio,
        self.box_noise_scale,
        self.training,
    )

    embed, refer_bbox, enc_bboxes, enc_scores = self._get_decoder_input(feats, shapes, dn_embed, dn_bbox)

    # Decoder
    dec_bboxes, dec_scores = self.decoder(
        embed,
        refer_bbox,
        feats,
        shapes,
        self.dec_bbox_head,
        self.dec_score_head,
        self.query_pos_head,
        attn_mask=attn_mask,
    )
    x = dec_bboxes, dec_scores, enc_bboxes, enc_scores, dn_meta
    if self.training:
        return x
    # (bs, 300, 4+nc)
    y = torch.cat((dec_bboxes.squeeze(0), dec_scores.squeeze(0).sigmoid()), -1)
    return y if self.export else (y, x)



ultralytics.nn.modules.head.v10Detect

├ťsler: Detect

v10 Detection head from https://arxiv.org/pdf/2405.14458

Parametreler:

─░sim Tip A├ž─▒klama Varsay─▒lan
nc int

S─▒n─▒f say─▒s─▒.

80
ch tuple

Tuple of channel sizes.

()

Nitelikler:

─░sim Tip A├ž─▒klama
max_det int

Maximum number of detections.

Y├Ântemler:

─░sim A├ž─▒klama
__init__

Initializes the v10Detect object.

forward

Performs forward pass of the v10Detect module.

bias_init

Initializes biases of the Detect module.

Kaynak kodu ultralytics/nn/modules/head.py
class v10Detect(Detect):
    """
    v10 Detection head from https://arxiv.org/pdf/2405.14458

    Args:
        nc (int): Number of classes.
        ch (tuple): Tuple of channel sizes.

    Attributes:
        max_det (int): Maximum number of detections.

    Methods:
        __init__(self, nc=80, ch=()): Initializes the v10Detect object.
        forward(self, x): Performs forward pass of the v10Detect module.
        bias_init(self): Initializes biases of the Detect module.

    """

    end2end = True

    def __init__(self, nc=80, ch=()):
        """Initializes the v10Detect object with the specified number of classes and input channels."""
        super().__init__(nc, ch)
        c3 = max(ch[0], min(self.nc, 100))  # channels
        # Light cls head
        self.cv3 = nn.ModuleList(
            nn.Sequential(
                nn.Sequential(Conv(x, x, 3, g=x), Conv(x, c3, 1)),
                nn.Sequential(Conv(c3, c3, 3, g=c3), Conv(c3, c3, 1)),
                nn.Conv2d(c3, self.nc, 1),
            )
            for x in ch
        )
        self.one2one_cv3 = copy.deepcopy(self.cv3)

__init__(nc=80, ch=())

Initializes the v10Detect object with the specified number of classes and input channels.

Kaynak kodu ultralytics/nn/modules/head.py
def __init__(self, nc=80, ch=()):
    """Initializes the v10Detect object with the specified number of classes and input channels."""
    super().__init__(nc, ch)
    c3 = max(ch[0], min(self.nc, 100))  # channels
    # Light cls head
    self.cv3 = nn.ModuleList(
        nn.Sequential(
            nn.Sequential(Conv(x, x, 3, g=x), Conv(x, c3, 1)),
            nn.Sequential(Conv(c3, c3, 3, g=c3), Conv(c3, c3, 1)),
            nn.Conv2d(c3, self.nc, 1),
        )
        for x in ch
    )
    self.one2one_cv3 = copy.deepcopy(self.cv3)





Created 2023-11-12, Updated 2024-06-20
Authors: Burhan-Q (2), glenn-jocher (7)