Skip to content

Perform Inference with trained network on novel videos #186

@WernTanj

Description

@WernTanj

I trained Resnet50 successfully with DLC on ground truth data. How can I perform inference with DLC using the best snapshot on a new, unseen video without creating a new training data set again? If I understand correctly the command 'deeplabcut.analyze_videos' requires the inputs 'config_path' and 'shuffle' which are created during the training process.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions