preprocess the dataset

[Joseph-LIN-Tech]

Thanks for your dataset. I found the folders in the downloaded data compression package looks messy.
Is there any scripts to process the data or is there a problem when packaging?

[seanko29]

@Joseph-LIN-Tech Did you solve the problem?
the HQ-EVFI Dataset has too much subfolders and there are no detailed explanation on how to solve them :(

[Enigmatic-bot-hash]

在主页有train和test，但是代码需要修改

[seanko29]

@Enigmatic-bot-hash yeah I saw that issue too that there are unrelated frames inserted within datasets.
Does that occur only when interp_ratio is set to 4?
I see in some other ratios. I didnt test this model in awhile.