TemporalLoss

class lightning_pose.losses.losses.TemporalLoss(data_module: BaseDataModule | UnlabeledDataModule | None = None, epsilon: float | List[float] = 0.0, prob_threshold: float = 0.0, log_weight: float = 0.0, **kwargs)

Bases: Loss

Penalize temporal differences for each target.

Motion model: x_t = x_(t-1) + e_t, e_t ~ N(0, s)

Methods Summary

__call__(keypoints_pred[, confidences, stage])	Call self as a function.
compute_loss(predictions)
rectify_epsilon(loss)	Rectify supporting a list of epsilons, one per bodypart.
remove_nans(loss, confidences)

Methods Documentation

__call__(keypoints_pred: Tensor[Tensor], confidences: Tensor[Tensor] | None = None, stage: Literal['train', 'val', 'test'] | None = None, **kwargs) → Tuple[Tensor[Tensor], List[dict]]

Call self as a function.

compute_loss(predictions: Tensor[Tensor]) → Tensor[Tensor]

rectify_epsilon(loss: Tensor[Tensor]) → Tensor[Tensor]

Rectify supporting a list of epsilons, one per bodypart. Not implemented in Loss class, because shapes of broadcasting may vary

remove_nans(loss: Tensor[Tensor], confidences: Tensor[Tensor]) → Tensor[Tensor]