Feature: support yolov10

configs/yolov10/hyp.scratch.yaml

@@ -0,0 +1,60 @@
+optimizer:
+  optimizer: momentum
+  lr_init: 0.01  # initial learning rate (SGD=1E-2, Adam=1E-3)
+  momentum: 0.937  # SGD momentum/Adam beta1
+  nesterov: True  # update gradients with NAG(Nesterov Accelerated Gradient) algorithm
+  loss_scale: 1.0  # loss scale for optimizer
+  warmup_epochs: 3  # warmup epochs (fractions ok)
+  warmup_momentum: 0.8  # warmup initial momentum
+  warmup_bias_lr: 0.1  # warmup initial bias lr
+  min_warmup_step: 1000  # minimum warmup step
+  group_param: yolov8  # group param strategy
+  gp_weight_decay: 0.0005  # group param weight decay 5e-4
+  start_factor: 1.0
+  end_factor: 0.01
+
+loss:
+  name: YOLOv10Loss
+  box: 7.5  # box loss gain
+  cls: 0.5  # cls loss gain
+  dfl: 1.5  # dfl loss gain
+  reg_max: 16
+
+data:
+  num_parallel_workers: 4
+
+  # multi-stage data augment
+  train_transforms: {
+    stage_epochs: [ 490, 10 ],
+    trans_list: [
+      [
+        {func_name: mosaic, prob: 1.0},
+        {func_name: resample_segments},
+        {func_name: random_perspective, prob: 1.0, degrees: 0.0, translate: 0.1, scale: 0.5, shear: 0.0},
+        {func_name: albumentations},
+        {func_name: hsv_augment, prob: 1.0, hgain: 0.015, sgain: 0.7, vgain: 0.4},
+        {func_name: fliplr, prob: 0.5},
+        {func_name: label_norm, xyxy2xywh_: True},
+        {func_name: label_pad, padding_size: 160, padding_value: -1},
+        {func_name: image_norm, scale: 255.},
+        {func_name: image_transpose, bgr2rgb: True, hwc2chw: True}
+      ],
+      [
+        {func_name: letterbox, scaleup: True},
+        {func_name: resample_segments},
+        {func_name: random_perspective, prob: 1.0, degrees: 0.0, translate: 0.1, scale: 0.5, shear: 0.0},
+        {func_name: albumentations},
+        {func_name: hsv_augment, prob: 1.0, hgain: 0.015, sgain: 0.7, vgain: 0.4},
+        {func_name: fliplr, prob: 0.5},
+        {func_name: label_norm, xyxy2xywh_: True},
+        {func_name: label_pad, padding_size: 160, padding_value: -1},
+        {func_name: image_norm, scale: 255.},
+        {func_name: image_transpose, bgr2rgb: True, hwc2chw: True}
+      ]]
+  }
+
+  test_transforms: [
+    {func_name: letterbox, scaleup: False, only_image: True},
+    {func_name: image_norm, scale: 255.},
+    {func_name: image_transpose, bgr2rgb: True, hwc2chw: True}
+  ]

configs/yolov10/yolov10n.yaml

@@ -0,0 +1,57 @@
+__BASE__: [
+  '../coco.yaml',
+  './hyp.scratch.yaml',
+]
+
+epochs: 500 # total train epochs
+per_batch_size: 16  # 16 * 8 = 128
+img_size: 640
+iou_thres: 0.7
+overflow_still_update: False
+opencv_threads_num: 0  # opencv: disable threading optimizations
+
+network:
+  model_name: yolov10
+  nc: 80  # number of classes
+  reg_max: 16
+
+  depth_multiple: 0.33  # model depth multiple
+  width_multiple: 0.25  # layer channel multiple
+  max_channels: 1024
+  stride: [8, 16, 32]
+
+  # YOLOv10.0n backbone
+  backbone:
+    # [from, repeats, module, args]
+    - [-1, 1, ConvNormAct, [64, 3, 2]]   # 0-P1/2
+    - [-1, 1, ConvNormAct, [128, 3, 2]]  # 1-P2/4
+    - [-1, 3, C2f, [128, True]]
+    - [-1, 1, ConvNormAct, [256, 3, 2]]  # 3-P3/8
+    - [-1, 6, C2f, [256, True]]
+    - [-1, 1, SCDown, [512, 3, 2]]       # 5-P4/16
+    - [-1, 6, C2f, [512, True]]
+    - [-1, 1, SCDown, [1024, 3, 2]]      # 7-P5/32
+    - [-1, 3, C2f, [1024, True]]
+    - [-1, 1, SPPF, [1024, 5]]           # 9
+    - [-1, 1, PSA, [1024, 5]]            # 10
+
+  # YOLOv10.0n head
+  head:
+    - [-1, 1, Upsample, [None, 2, 'nearest']]
+    - [[-1, 6], 1, Concat, [1]]  # cat backbone P4
+    - [-1, 3, C2f, [512]]  # 13
+
+    - [-1, 1, Upsample, [None, 2, 'nearest']]
+    - [[-1, 4], 1, Concat, [1] ]  # cat backbone P3
+    - [-1, 3, C2f, [256]]  # 16 (P3/8-small)
+
+    - [-1, 1, ConvNormAct, [256, 3, 2]]
+    - [[ -1, 13], 1, Concat, [1]]  # cat head P4
+    - [-1, 3, C2f, [512]]  # 19 (P4/16-medium)
+
+    - [-1, 1, SCDown, [512, 3, 2]]
+    - [[-1, 10], 1, Concat, [1]]  # cat head P5
+    - [-1, 3, C2fCIB, [1024, True, True]]  # 22 (P5/32-large)
+
+    - [[16, 19, 22], 1, YOLOv10Head, [nc, reg_max, stride]]  # Detect(P3, P4, P5)
+

mindyolo/models/__init__.py

@@ -1,10 +1,11 @@
 from . import (heads, initializer, layers, losses, model_factory, yolov3,
-               yolov4, yolov5, yolov7, yolov8)
+               yolov4, yolov5, yolov7, yolov8, yolov10)
 
 __all__ = []
 __all__.extend(heads.__all__)
 __all__.extend(layers.__all__)
 __all__.extend(losses.__all__)
+__all__.extend(yolov10.__all__)
 __all__.extend(yolov8.__all__)
 __all__.extend(yolov7.__all__)
 __all__.extend(yolov5.__all__)
@@ -25,4 +26,5 @@
 from .yolov5 import *
 from .yolov7 import *
 from .yolov8 import *
+from .yolov10 import *
 from .yolox import *

mindyolo/models/heads/__init__.py

@@ -5,13 +5,14 @@
 from .yolov7_head import *
 from .yolov8_head import *
 from .yolox_head import *
-
+from .yolov10_head import *
 
 __all__ = [
     "YOLOv3Head",
     "YOLOv4Head",
     "YOLOv5Head",
     "YOLOv7Head", "YOLOv7AuxHead",
     "YOLOv8Head", "YOLOv8SegHead",
-    "YOLOXHead"
+    "YOLOXHead",
+    "YOLOv10Head"
 ]

mindyolo/models/heads/yolov10_head.py

@@ -0,0 +1,194 @@
+import math
+import numpy as np
+from copy import deepcopy
+
+import mindspore as ms
+import mindspore.numpy as mnp
+from mindspore import Parameter, Tensor, nn, ops
+
+from ..layers import DFL, ConvNormAct, Identity
+from ..layers.utils import meshgrid
+
+class YOLOv10Head(nn.Cell):
+    # YOLOv10 Detect head for detection models
+    def __init__(self, nc=80, reg_max=16, stride=(), ch=(), sync_bn=False):  # detection layer
+        super().__init__()
+        # self.dynamic = False # force grid reconstruction
+
+        assert isinstance(stride, (tuple, list)) and len(stride) > 0
+        assert isinstance(ch, (tuple, list)) and len(ch) > 0
+
+        self.nc = nc  # number of classes
+        self.nl = len(ch)  # number of detection layers
+        self.reg_max = reg_max  # 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 = Parameter(Tensor(stride, ms.int32), requires_grad=False)
+        self.max_det = 300  # max_det
+        self.end2end = True
+        self.export = False
+
+        c2, c3 = max((16, ch[0] // 4, self.reg_max * 4)), max(ch[0], min(self.nc, 100))  # channels
+        self.cv2 = nn.CellList(
+            [
+                nn.SequentialCell(
+                    [
+                        ConvNormAct(x, c2, 3, sync_bn=sync_bn),
+                        ConvNormAct(c2, c2, 3, sync_bn=sync_bn),
+                        nn.Conv2d(c2, 4 * self.reg_max, 1, has_bias=True),
+                    ]
+                )
+                for x in ch
+            ]
+        )
+        self.cv3 = nn.CellList(
+            [
+                nn.SequentialCell(
+                    [
+                        nn.SequentialCell(
+                            [
+                                ConvNormAct(x, x, 3, g=x),
+                                ConvNormAct(x, c3, 1)
+                            ]
+                        ),
+                        nn.SequentialCell([
+                                ConvNormAct(c3, c3, 3, g=c3),
+                                ConvNormAct(c3, c3, 1)
+                            ]
+                        ),
+                        nn.Conv2d(c3, self.nc, 1, has_bias=True)
+                    ]
+                )
+                for i, x in enumerate(ch)
+            ]
+        )
+        self.dfl = DFL(self.reg_max) if self.reg_max > 1 else Identity()
+
+        self.one2one_cv2 = deepcopy(self.cv2)
+        self.one2one_cv3 = deepcopy(self.cv3)
+
+    def construct(self, x):
+        """Concatenates and returns predicted bounding boxes and class probabilities."""
+        if self.end2end:
+            return self.construct_end2end(x)
+
+        x = ()
+        for i in range(self.nl):
+            x += (ops.concat((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 construct_end2end(self, x):
+        """
+        Performs forward pass of the YOLOv10Head 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 = [ops.stop_gradient(xi) for xi in x]
+        one2one = ()
+        for i in range(self.nl):
+            one2one += (ops.concat((self.one2one_cv2[i](x_detach[i]), self.one2one_cv3[i](x_detach[i])), 1),)
+        one2many = ()
+        for i in range(self.nl):
+            one2many += (ops.concat((self.cv2[i](x[i]), self.cv3[i](x[i])), 1), )
+        if self.training:  # Training path
+            return (one2many, one2one)
+        y = self._inference(one2one)
+        y = self.postprocess(y, self.max_det, self.nc)
+        return y if self.export else (y, {"one2many": one2many, "one2one": one2one})
+
+    def _inference(self, x):
+        # Inference path
+        shape = x[0].shape  # BCHW
+        _anchors, _strides = self.make_anchors(x, self.stride, 0.5)
+        _anchors, _strides = _anchors.swapaxes(0, 1), _strides.swapaxes(0, 1)
+
+        _x = ()
+        for i in range(len(x)):
+            _x += (x[i].view(shape[0], self.no, -1),)
+        _x = ops.concat(_x, 2)
+        box, cls = _x[:, : self.reg_max * 4, :], _x[:, self.reg_max * 4 : self.reg_max * 4 + self.nc, :]
+        # box, cls = ops.concat([xi.view(shape[0], self.no, -1) for xi in x], 2).split((self.reg_max * 4, self.nc), 1)
+        dbox = self.dist2bbox(self.dfl(box), ops.expand_dims(_anchors, 0), xywh=not self.end2end, axis=1) * _strides
+
+        y = None
+        y = ops.concat((dbox, ops.Sigmoid()(cls)), 1)
+        y = ops.transpose(y, (0, 2, 1))  # (bs, no-84, nbox) -> (bs, nbox, no-84)
+        return (y, x)
+
+    @staticmethod
+    def make_anchors(feats, strides, grid_cell_offset=0.5):
+        """Generate anchors from features."""
+        anchor_points, stride_tensor = (), ()
+        dtype = feats[0].dtype
+        for i, stride in enumerate(strides):
+            _, _, h, w = feats[i].shape
+            sx = mnp.arange(w, dtype=dtype) + grid_cell_offset  # shift x
+            sy = mnp.arange(h, dtype=dtype) + grid_cell_offset  # shift y
+            # FIXME: Not supported on a specific model of machine
+            sy, sx = meshgrid((sy, sx), indexing="ij")
+            anchor_points += (ops.stack((sx, sy), -1).view(-1, 2),)
+            stride_tensor += (ops.ones((h * w, 1), dtype) * stride,)
+        return ops.concat(anchor_points), ops.concat(stride_tensor)
+
+    @staticmethod
+    def dist2bbox(distance, anchor_points, xywh=True, axis=-1):
+        """Transform distance(ltrb) to box(xywh or xyxy)."""
+        lt, rb = ops.split(distance, split_size_or_sections=2, axis=axis)
+        x1y1 = anchor_points - lt
+        x2y2 = anchor_points + rb
+        if xywh:
+            c_xy = (x1y1 + x2y2) / 2
+            wh = x2y2 - x1y1
+            return ops.concat((c_xy, wh), axis)  # xywh bbox
+        return ops.concat((x1y1, x2y2), axis)  # xyxy bbox
+
+    @staticmethod
+    def postprocess(preds, max_det, nc=80):
+        """
+        Post-processes YOLO model predictions.
+
+        Args:
+            preds (torch.Tensor): Raw predictions with shape (batch_size, num_anchors, 4 + nc) with last dimension
+                format [x, y, w, h, class_probs].
+            max_det (int): Maximum detections per image.
+            nc (int, optional): Number of classes. Default: 80.
+
+        Returns:
+            (torch.Tensor): Processed predictions with shape (batch_size, min(max_det, num_anchors), 6) and last
+                dimension format [x, y, w, h, max_class_prob, class_index].
+        """
+        batch_size, _, _ = preds.shape  # i.e. shape(16,8400,84)
+        boxes, scores = preds.split([4, nc], dim=-1)
+        max_scores = ops.amax(scores, dim=-1)
+        max_scores, index = ops.topk(max_scores, max_det, dim=-1)
+        index = ops.expand_dims(index, -1)
+        boxes = ops.gather(boxes, dim=1, index=index.repeat(1, 1, boxes.shape[-1]))
+        scores = ops.gather(scores, dim=1, index=index.repeat(1, 1, nc))
+
+        scores, index = ops.topk(ops.flatten(scores, start_dim=1), max_det, dim=-1)
+        i = ops.arange(batch_size)[..., None]  # batch indices
+        return ops.cat([boxes[i, index // nc], scores[..., None], (index % nc)[..., None].float()], dim=-1)
+
+    def initialize_biases(self):
+        # Initialize Detect() biases, WARNING: requires stride availability
+        m = self
+        for a, b, s in zip(m.cv2, m.cv3, m.stride):  # from
+            s = s.asnumpy()
+            a[-1].bias = ops.assign(a[-1].bias, Tensor(np.ones(a[-1].bias.shape), ms.float32))
+            b_np = b[-1].bias.data.asnumpy()
+            b_np[: m.nc] = math.log(5 / m.nc / (640 / int(s)) ** 2)
+            b[-1].bias = ops.assign(b[-1].bias, Tensor(b_np, ms.float32))
+        if self.end2end:
+            for a, b, s in zip(m.one2one_cv2, m.one2one_cv3, m.stride):  # from
+                s = s.asnumpy()
+                a[-1].bias = ops.assign(a[-1].bias, Tensor(np.ones(a[-1].bias.shape), ms.float32))
+                b_np = b[-1].bias.data.asnumpy()
+                b_np[: m.nc] = math.log(5 / m.nc / (640 / int(s)) ** 2)
+                b[-1].bias = ops.assign(b[-1].bias, Tensor(b_np, ms.float32))

mindyolo/models/layers/__init__.py

@@ -34,4 +34,7 @@
     "SPPF",
     "Upsample",
     "Residualblock",
+    "SCDown",
+    "PSA",
+    "C2fCIB",
 ]