Skip to content

Image normalization for feature extraction #44

Description

@despotovic

First of all, thank you very much for the excellent work!

When using CTransPath for feature extraction, you normalize with the following image statistics:
mean = (0.485, 0.456, 0.406)
std = (0.229, 0.224, 0.225)

Are this mean and standard deviation of the dataset used for pretraining the model? Should I adapt this with the mean and standard deviation of my own dataset when doing feature extraction?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions