bbox_to_activ in metrics.py

[mbuttery-ghd]

walkwithfastai.github.io/wwf/vision/object_detection/metrics.py

Lines 17 to 24 in ecdd5ed

	def bbox_to_activ(bboxes, anchors, flatten=True):
	"Return the target of the model on `anchors` for the `bboxes`."
	if flatten:
	t_centers = (bboxes[...,:2] - anchors[...,:2]) / anchors[...,2:]
	t_sizes = torch.log(bboxes[...,2:] / anchors[...,2:] + 1e-8)
	return torch.cat([t_centers, t_sizes], -1).div_(bboxes.new_tensor([[0.1, 0.1, 0.2, 0.2]]))
	else: return [activ_to_bbox(act,anc) for act,anc in zip(acts, anchors)]
	return res

Shouldn't line 23 be

else: return [bbox_to_activ(bbox, anc) for bbox, anc in zip(bboxes, anchors)]