Skip to content

Question about Horse2Zebra dataset size discrepancy #161

Description

@chaksseu

Hello,
Thank you for sharing your great work.

While trying to reproduce the Horse2Zebra experiments, I noticed a discrepancy between the dataset sizes reported in the paper and the dataset available for download.

In the paper, the reported numbers are:
TrainA(horse): 939 images
TrainB(zebra): 1177 images
However, the dataset I downloaded contains:
TrainA(horse): 1067 images
TrainB(zebra): 1334 images

Could you clarify whether:
A different version of the dataset was used (e.g., filtered or preprocessed)?
Some samples were excluded during training?
The exact dataset split used in the paper is available?

I would appreciate any clarification, as I am trying to reproduce the results as faithfully as possible.

Thank you in advance!

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