PointCloudObjectDetector

class flash.pointcloud.detection.model.PointCloudObjectDetector(num_classes, backbone='pointpillars_kitti', backbone_kwargs=None, loss_fn=None, optimizer='Adam', lr_scheduler=None, metrics=None, learning_rate=None, lambda_loss_cls=1.0, lambda_loss_bbox=1.0, lambda_loss_dir=1.0)

The PointCloudObjectDetector is a ClassificationTask that classifies pointcloud data.

Parameters

• num_classes (int) – The number of classes (outputs) for this Task.

• backbone (Union[str, Tuple[Module, int]]) – The backbone name (or a tuple of nn.Module, output size) to use.

• backbone_kwargs (Optional[Dict]) – Any additional kwargs to pass to the backbone constructor.

• loss_fn (Optional[TypeVar(LOSS_FN_TYPE, Callable, Mapping, Sequence, None)]) – The loss function to use. If None, a default will be selected by the ClassificationTask depending on the multi_label argument.

• optimizer (TypeVar(OPTIMIZER_TYPE, str, Callable, Tuple[str, Dict[str, Any]], None)) – Optimizer to use for training.

• lr_scheduler (Optional[TypeVar(LR_SCHEDULER_TYPE, str, Callable, Tuple[str, Dict[str, Any]], Tuple[str, Dict[str, Any], Dict[str, Any]], None)]) – The LR scheduler to use during training.

• metrics (Optional[TypeVar(METRICS_TYPE, Metric, Mapping, Sequence, None)]) – Any metrics to use with this Task. If None, a default will be selected by the ClassificationTask depending on the multi_label argument.

• learning_rate (Optional[float]) – The learning rate for the optimizer.

• lambda_loss_cls (float) – The value to scale the loss classification.

• lambda_loss_bbox (float) – The value to scale the bounding boxes loss.

• lambda_loss_dir (float) – The value to scale the bounding boxes direction loss.

classmethod available_finetuning_strategies(cls)

Returns a list containing the keys of the available Finetuning Strategies.

Return type

List[str]

classmethod available_lr_schedulers(cls)

Returns a list containing the keys of the available LR schedulers.

Return type

List[str]

classmethod available_optimizers(cls)

Returns a list containing the keys of the available Optimizers.

Return type

List[str]

classmethod available_outputs(cls)

Returns the list of available outputs (that can be used during prediction or serving) for this Task.

Examples

..testsetup:

>>> from flash import Task

>>> print(Task.available_outputs())
['preds', 'raw']

Return type

List[str]

forward(x)

First call the backbone, then the model head.

Return type

Tensor