Missing fine-grained aesthetic dimension labels
#1
by karsish - opened
Hi,
Thank you for releasing VideoAesBench. I noticed that videoaesbench_qa.jsonl provides the three main aesthetic dimensions (visual form, visual style, and visual affectiveness), but I could not find the labels for the 12 fine-grained aesthetic dimensions described in the paper.
Would it be possible to release these fine-grained labels for each question? They would be helpful for reproducing the dimension-level evaluation reported in the paper.
Thank you!