- class mmcv.ops.RoIAlignRotated(output_size: Union[int, tuple], spatial_scale: float, sampling_ratio: int = 0, aligned: bool = True, clockwise: bool = False)¶
RoI align pooling layer for rotated proposals.
It accepts a feature map of shape (N, C, H, W) and rois with shape (n, 6) with each roi decoded as (batch_index, center_x, center_y, w, h, angle). The angle is in radian.
output_size (tuple) – h, w
spatial_scale (float) – scale the input boxes by this number
sampling_ratio (int) – number of inputs samples to take for each output sample. 0 to take samples densely for current models.
aligned (bool) – if False, use the legacy implementation in MMDetection. If True, align the results more perfectly. Default: True.
clockwise (bool) – If True, the angle in each proposal follows a clockwise fashion in image space, otherwise, the angle is counterclockwise. Default: False.
The implementation of RoIAlign when aligned=True is modified from https://github.com/facebookresearch/detectron2/
The meaning of aligned=True:
Given a continuous coordinate c, its two neighboring pixel indices (in our pixel model) are computed by floor(c - 0.5) and ceil(c - 0.5). For example, c=1.3 has pixel neighbors with discrete indices  and  (which are sampled from the underlying signal at continuous coordinates 0.5 and 1.5). But the original roi_align (aligned=False) does not subtract the 0.5 when computing neighboring pixel indices and therefore it uses pixels with a slightly incorrect alignment (relative to our pixel model) when performing bilinear interpolation.
With aligned=True, we first appropriately scale the ROI and then shift it by -0.5 prior to calling roi_align. This produces the correct neighbors;
The difference does not make a difference to the model’s performance if ROIAlign is used together with conv layers.
- forward(input: torch.Tensor, rois: torch.Tensor) → torch.Tensor¶
Defines the computation performed at every call.
Should be overridden by all subclasses.
Although the recipe for forward pass needs to be defined within this function, one should call the
Moduleinstance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.