data_pipeline.md

Tutorial 2: How to use Data Pipelines

In this tutorial, we concept of data pipelines, and how to customize and extend your own data pipelines for the project.

• Tutorial 2: How to use Data Pipelines
  • Concept of Data Pipelines
  • Design of Data Pipelines

Concept of Data Pipelines

Data Pipeline is a modular form for data process. We make common data processing operations into python class, which named pipeline. For example, in image tasks, pre-processing often involves cropping, deformation, color, adding noise, etc. It is named ````pipeline``` in the mmcv series of codes.

The preprocessing flow is defined in the config file:

train_pipeline = [
    dict(type='ToTensor', enable=True, keys=['music', 'dance'],),
]

For dance generation movements, since there is no pre-processing of music or action sequences in the current general algorithms, there is only one process - ToTensor - which converts music/dance sequences into torch.tensor.

Pipelines are usually defined under the datasets folder. A template for a custom pipeline is as follows.

@PIPELINES.register_module()
class PipelineA:
    """get viewdirs from rays_d
    """
    def __init__(self, enable=True, **kwargs):
        self.enable = enable

    def __call__(self, results):
        """get viewdirs
        Args:
            results (dict): The resulting dict to be modified and passed
                to the next transform in pipeline.
        """
        # process on results
        return results