RTMDet-tiny enablement for detection task (#3542)

- RTMDet-tiny enablement on detection task
- Minor refactoring

src/otx/algo/detection/atss.py

@@ -104,7 +104,7 @@ def _customize_outputs(
                 elif isinstance(v, torch.Tensor):
                     losses[k] = v
                 else:
-                    msg = "Loss output should be list or torch.tensor but got {type(v)}"
+                    msg = f"Loss output should be list or torch.tensor but got {type(v)}"
                     raise TypeError(msg)
             return losses
 

src/otx/algo/detection/heads/atss_head.py

@@ -12,6 +12,7 @@
 from otx.algo.detection.heads.class_incremental_mixin import (
     ClassIncrementalMixin,
 )
+from otx.algo.detection.losses.cross_entropy_loss import CrossEntropyLoss
 from otx.algo.detection.losses.cross_focal_loss import (
     CrossSigmoidFocalLoss,
 )
@@ -57,12 +58,12 @@ def __init__(
         self,
         num_classes: int,
         in_channels: int,
-        loss_centerness: nn.Module,
         pred_kernel_size: int = 3,
         stacked_convs: int = 4,
         conv_cfg: dict | None = None,
         norm_cfg: dict | None = None,
         reg_decoded_bbox: bool = True,
+        loss_centerness: nn.Module | None = None,
         init_cfg: dict | None = None,
         bg_loss_weight: float = -1.0,
         use_qfl: bool = False,
@@ -89,7 +90,7 @@ def __init__(
         )
 
         self.sampling = False
-        self.loss_centerness = loss_centerness
+        self.loss_centerness = loss_centerness or CrossEntropyLoss(use_sigmoid=True, loss_weight=1.0)
 
         if use_qfl:
             kwargs["loss_cls"] = (

src/otx/algo/detection/heads/delta_xywh_bbox_coder.py

@@ -8,6 +8,8 @@
 import torch
 from torch import Tensor
 
+from otx.algo.detection.utils.utils import clip_bboxes_export
+
 
 # This class and its supporting functions below lightly adapted from the mmdet DeltaXYWHBBoxCoder available at:
 # https://github.com/open-mmlab/mmdetection/blob/cfd5d3a985b0249de009b67d04f37263e11cdf3d/mmdet/models/task_modules/coders/delta_xywh_bbox_coder.py
@@ -360,54 +362,6 @@ def delta2bbox_export(
     y2 = xy2[..., 1]
 
     if clip_border and max_shape is not None:
-        x1, y1, x2, y2 = clip_bboxes(x1, y1, x2, y2, max_shape)
+        x1, y1, x2, y2 = clip_bboxes_export(x1, y1, x2, y2, max_shape)
 
     return torch.stack([x1, y1, x2, y2], dim=-1).view(deltas.size())
-
-
-def clip_bboxes(
-    x1: Tensor,
-    y1: Tensor,
-    x2: Tensor,
-    y2: Tensor,
-    max_shape: Tensor | tuple[int, ...],
-) -> tuple[Tensor, ...]:
-    """Clip bboxes for onnx.
-
-    Since torch.clamp cannot have dynamic `min` and `max`, we scale the
-      boxes by 1/max_shape and clamp in the range [0, 1] if necessary.
-
-    Args:
-        x1 (Tensor): The x1 for bounding boxes.
-        y1 (Tensor): The y1 for bounding boxes.
-        x2 (Tensor): The x2 for bounding boxes.
-        y2 (Tensor): The y2 for bounding boxes.
-        max_shape (Tensor | Sequence[int]): The (H,W) of original image.
-
-    Returns:
-        tuple(Tensor): The clipped x1, y1, x2, y2.
-    """
-    if isinstance(max_shape, torch.Tensor):
-        # scale by 1/max_shape
-        x1 = x1 / max_shape[1]
-        y1 = y1 / max_shape[0]
-        x2 = x2 / max_shape[1]
-        y2 = y2 / max_shape[0]
-
-        # clamp [0, 1]
-        x1 = torch.clamp(x1, 0, 1)
-        y1 = torch.clamp(y1, 0, 1)
-        x2 = torch.clamp(x2, 0, 1)
-        y2 = torch.clamp(y2, 0, 1)
-
-        # scale back
-        x1 = x1 * max_shape[1]
-        y1 = y1 * max_shape[0]
-        x2 = x2 * max_shape[1]
-        y2 = y2 * max_shape[0]
-    else:
-        x1 = torch.clamp(x1, 0, max_shape[1])
-        y1 = torch.clamp(y1, 0, max_shape[0])
-        x2 = torch.clamp(x2, 0, max_shape[1])
-        y2 = torch.clamp(y2, 0, max_shape[0])
-    return x1, y1, x2, y2

src/otx/algo/detection/heads/distance_point_bbox_coder.py

@@ -1,13 +1,13 @@
 # Copyright (C) 2024 Intel Corporation
 # SPDX-License-Identifier: Apache-2.0
 # Copyright (c) OpenMMLab. All rights reserved.
-""""Distance Point BBox coder."""
+"""Distance Point BBox coder."""
 
 from __future__ import annotations
 
 from typing import TYPE_CHECKING
 
-from otx.algo.detection.utils.utils import bbox2distance, distance2bbox
+from otx.algo.detection.utils.utils import bbox2distance, distance2bbox, distance2bbox_export
 
 if TYPE_CHECKING:
     from torch import Tensor
@@ -84,3 +84,30 @@ def decode(
         if self.clip_border is False:
             max_shape = None
         return distance2bbox(points, pred_bboxes, max_shape)
+
+    def decode_export(
+        self,
+        points: Tensor,
+        pred_bboxes: Tensor,
+        max_shape: tuple[int, ...] | Tensor | tuple[tuple[int, ...], ...] | None = None,
+    ) -> Tensor:
+        """Decode distance prediction to bounding box for export."""
+        if points.size(0) != pred_bboxes.size(0):
+            msg = (
+                f"The batch of points (={points.size(0)}) and the batch of pred_bboxes "
+                f"(={pred_bboxes.size(0)}) should be same."
+            )
+            raise ValueError(msg)
+
+        if points.size(-1) != 2:
+            msg = f"points should have the format with size of 2, given {points.size(-1)}."
+            raise ValueError(msg)
+
+        if pred_bboxes.size(-1) != 4:
+            msg = f"pred_bboxes should have the format with size of 4, given {pred_bboxes.size(-1)}."
+            raise ValueError(msg)
+
+        if self.clip_border is False:
+            max_shape = None
+
+        return distance2bbox_export(points, pred_bboxes, max_shape)