mindformers.pipeline.ImageClassificationPipeline

class mindformers.pipeline.ImageClassificationPipeline(model: Union[str, BaseModel, Model], image_processor: Optional[BaseImageProcessor] = None, **kwargs)

Pipeline for image classification

Args:

model (Union[str, BaseModel]): The model used to perform task,

the input could be a supported model name, or a model instance inherited from BaseModel.

image_processor (Optional[BaseImageProcessor]): The image_processor of model,

it could be None if the model do not need image_processor.

Raises:

TypeError: If input model and image_processor’s types are not corrected. ValueError: If the input model is not in support list.

Examples:

>>> import numpy as np
>>> from mindformers.pipeline import ImageClassificationPipeline
>>> from mindformers import ViTImageProcessor
>>> processor = ViTImageProcessor(size=224)
>>> classifier = ImageClassificationPipeline(
...     model='vit_base_p16',
...     image_processor=processor,
...     top_k=5
...     )
>>> classifier(np.uint8(np.random.random((5, 3, 255, 255))))
    [[{'score': 0.0016654134, 'label': 'matchstick'},
    {'score': 0.0015071577, 'label': 'theater curtain'},
    {'score': 0.0014839625, 'label': 'ocarina'},
    {'score': 0.0014319294, 'label': 'abaya'},
    {'score': 0.0014109017, 'label': 'bottlecap'}],
    ..., {'score': 0.0014109018, 'label': 'bottlecap'}]]

forward(model_inputs: dict, **forward_params)

The Forward Process of Model

Args:

model_inputs (dict): The output of preprocess. forward_params (dict): The parameter dict for model forward.

postprocess(model_outputs, **postprocess_params)

Postprocess

Args:

model_outputs (dict): Outputs of forward process. top_k (int): Return top_k probs of result.

Return:

classification results.

preprocess(inputs: Union[str, Image, Tensor, ndarray], **preprocess_params)

The Preprocess For Task

Args:

inputs (Union[url, PIL.Image, tensor, numpy]): The image to be classified. preprocess_params (dict): The parameter dict for preprocess.

Return:

Processed image.