A Review of Human Evaluation for Style Transfer
This addresses the problem of inconsistent evaluation practices in style transfer research, which is incremental as it summarizes existing issues without proposing new solutions.
The paper reviewed human evaluation practices in 97 style transfer papers, finding that protocols are often underspecified and not standardized, which hampers reproducibility and progress in evaluation methods.
This paper reviews and summarizes human evaluation practices described in 97 style transfer papers with respect to three main evaluation aspects: style transfer, meaning preservation, and fluency. In principle, evaluations by human raters should be the most reliable. However, in style transfer papers, we find that protocols for human evaluations are often underspecified and not standardized, which hampers the reproducibility of research in this field and progress toward better human and automatic evaluation methods.