Hello,
I have a question about an experiment in my selected paper `Image-to-image Translation via Hierarchical Style Disentanglement`.
I realized that in one part of an experiment, they measured Diversity by conducting a user study instead of using a metric like learned perceptual image patch similarity.
In other parts of this experiment and other ablation studies, they either use quantitative metrics like FID as usual, or have qualitative results.
Will I have to perform this part, or would it affect my grade negatively? In that case, would it be possible for me to switch to the second paper in my list (assuming the selections haven't been finalized yet).
Thanks,
Yours sincerely,
Caner Korkmaz