YouCook2-Leaderboard
YouCook2-Leaderboard copied to clipboard
add Nishimura et al. ACMMM21
I added my ACMMM21 work to the leaderboard.
@misogil0116 Thanks! Just out of curiosity, I notice the baseline numbers reported in your Tab. 2 (e.g., MART, Transformer-XL) are slightly different from the original paper. Could you elaborate a bit on what are the causes (e.g., the discrepancy in data)? Thanks.
@LuoweiZhou The difference of the scores are caused by the dataset split. In 2021 winter, I could not access the original test set annotations, so split the val set with val/test set, and evaluate the performance on the test set (I think that "DPC" described in the Leaderboard also uses their split). I will add notes to the commits.