mmcls.visualization

该包中包含了用于分类任务的一些可视化组件。

ClsVisualizer

class mmcls.visualization.ClsVisualizer(name='visualizer', image=None, vis_backends=None, save_dir=None, fig_save_cfg={'frameon': False}, fig_show_cfg={'frameon': False})

Universal Visualizer for classification task.

参数

• name (str) – Name of the instance. Defaults to ‘visualizer’.

• image (np.ndarray, optional) – the origin image to draw. The format should be RGB. Defaults to None.

• vis_backends (list, optional) – Visual backend config list. Defaults to None.

• save_dir (str, optional) – Save file dir for all storage backends. If it is None, the backend storage will not save any data.

• fig_save_cfg (dict) – Keyword parameters of figure for saving. Defaults to empty dict.

• fig_show_cfg (dict) – Keyword parameters of figure for showing. Defaults to empty dict.

使用示例

>>> import torch
>>> import mmcv
>>> from pathlib import Path
>>> from mmcls.visualization import ClsVisualizer
>>> from mmcls.structures import ClsDataSample
>>> # Example image
>>> img = mmcv.imread("./demo/bird.JPEG", channel_order='rgb')
>>> # Example annotation
>>> data_sample = ClsDataSample().set_gt_label(1).set_pred_label(1).        ...     set_pred_score(torch.tensor([0.1, 0.8, 0.1]))
>>> # Setup the visualizer
>>> vis = ClsVisualizer(
...     save_dir="./outputs",
...     vis_backends=[dict(type='LocalVisBackend')])
>>> # Set classes names
>>> vis.dataset_meta = {'classes': ['cat', 'bird', 'dog']}
>>> # Show the example image with annotation in a figure.
>>> # And it will ignore all preset storage backends.
>>> vis.add_datasample('res', img, data_sample, show=True)
>>> # Save the visualization result by the specified storage backends.
>>> vis.add_datasample('res', img, data_sample)
>>> assert Path('./outputs/vis_data/vis_image/res_0.png').exists()
>>> # Save another visualization result with the same name.
>>> vis.add_datasample('res', img, data_sample, step=1)
>>> assert Path('./outputs/vis_data/vis_image/res_1.png').exists()