How to run inferience for a single image (not included in the training set)

[sjcotto]

Hey, great project

How to run inferience for a single image (not included in the training set)

Team have plans for a google colab?

[ykk648]

working on it https://github.com/sangyun884/HR-VITON/issues/45

[ABaldrati]

Hi @sjcotto Thank you for reaching out and showing interest in our work! Regarding your request for a Google Colab demonstrating the specific functionality you mentioned, we currently do not have plans to release such a Colab.

To run inference on an image, you will need the following components:

• Model image: This is the image you want to use as a reference for the desired output.
• In-shop garment image: This is the image of the clothing item you want to transfer onto the model image.-
• Model image pose: Our system utilizes 18 key points to represent the pose of the model in the image.
• Model image label map: This is a label map specific to the model image, indicating different regions of the body.

To perform inference, you would need to process the model image using the same code as our custom PyTorch dataset. Once you have processed the model image with these ingredients, you should be able to run the inference successfully.

For a guide on starting this process, @ykk648 has provided a useful resource that you can refer to.

Alberto