Training on custom dataset with custom mask for Inpainting

[Akshay-Ijantkar]

Hello @lindahua @wwhio @tpoisonooo @hellock @innerlee
Great work on the paper and even the results look amazing!

Can you please guide me with step by step process on how we need to prepare a dataset for a custom image dataset with a custom mask (if possible)?

How the masks will look like in the sense that whether it will be a colour image on which white patches (the part that will be inpainted from the model) is drawn or it will be a completely black image on which white patches (the part that will be inpainted from the model) is drawn? (in short how to prepare mask and then how to supply it for model training)

I referred to your steps but couldn't completely follow them.
Thank you in advance!

[wwhio]

1. Implement your dataset which gives the information of your data. Register it, import it in mmedit/datasets/__init__.py, and add it to __all__.
2. Implement your pipeline which gives the process according to information provided by dataset. Register it, import it in mmedit/datasets/pipelines/__init__.py, and add it to __all__.

During implementing your dataset, just give the information:

1. Your dataset should inherit the ImgInpaintingDataset class, and implement your load_annotations function.
2. You may change the __init__ function according to your need.
3. In load_annotations: package all information you need for one paired data into a dict and append these dicts to a list and return this list.
4. Call load_annotations in __init__ like ImgInpaintingDataset does.

[ACALJJ32]

1. Implement your dataset which gives the information of your data. Register it, import it in mmedit/datasets/__init__.py, and add it to __all__.
2. Implement your pipeline which gives the process according to information provided by dataset. Register it, import it in mmedit/datasets/pipelines/__init__.py, and add it to __all__.

During implementing your dataset, just give the information:

1. Your dataset should inherit the ImgInpaintingDataset class, and implement your load_annotations function.
2. You may change the __init__ function according to your need.
3. In load_annotations: package all information you need for one paired data into a dict and append these dicts to a list and return this list.
4. Call load_annotations in __init__ like ImgInpaintingDataset does.

Thank you for this guide,it help me a lot!