MultiLabelClsHead

class mmpretrain.models.heads.MultiLabelClsHead(loss={'type': 'CrossEntropyLoss', 'use_sigmoid': True}, thr=None, topk=None, init_cfg=None)

Classification head for multilabel task.

Parameters:

• loss (dict) – Config of classification loss. Defaults to dict(type=’CrossEntropyLoss’, use_sigmoid=True).

• thr (float, optional) – Predictions with scores under the thresholds are considered as negative. Defaults to None.

• topk (int, optional) – Predictions with the k-th highest scores are considered as positive. Defaults to None.

• init_cfg (dict, optional) – The extra init config of layers. Defaults to None.

Notes

If both thr and topk are set, use thr` to determine positive predictions. If neither is set, use ``thr=0.5 as default.

forward(feats)

The forward process.

loss(feats, data_samples, **kwargs)

Calculate losses from the classification score.

Parameters:

• feats (tuple[Tensor]) – The features extracted from the backbone. Multiple stage inputs are acceptable but only the last stage will be used to classify. The shape of every item should be (num_samples, num_classes).

• data_samples (List[DataSample]) – The annotation data of every samples.

• **kwargs – Other keyword arguments to forward the loss module.

Returns:

a dictionary of loss components

Return type:

dict[str, Tensor]

pre_logits(feats)

The process before the final classification head.

The input feats is a tuple of tensor, and each tensor is the feature of a backbone stage. In MultiLabelClsHead, we just obtain the feature of the last stage.

predict(feats, data_samples=None)

Inference without augmentation.

Parameters:

• feats (tuple[Tensor]) – The features extracted from the backbone. Multiple stage inputs are acceptable but only the last stage will be used to classify. The shape of every item should be (num_samples, num_classes).

• data_samples (List[DataSample], optional) – The annotation data of every samples. If not None, set pred_label of the input data samples. Defaults to None.

Returns:

A list of data samples which contains the predicted results.

Return type:

List[DataSample]