Output dataset of taichi can not found
JialeTao opened this issue · 4 comments
Hi, I follow the command in the picture to pre-process taichi dataset and it is strangely the output folder is empty. The folder is created while there is no file in the folder. No errors assert in the running time. Does that mean the processed images are not saved in the code? Do you have any suggessions?
Take a digression, generally there are some package errors when directly run the preprocesing codes for taichi dataset. And it's mainly caused by the maskrcnn-benchmark building. I suggest to add a statement of asking followers to build maskrcnn-benchmark first.
Most likely you are running the script on the already cropped videos. Can you verify this? And if so, why are you doing this?
Most likely you are running the script on the already cropped videos. Can you verify this? And if so, why are you doing this?
Hi, thanks for quick reply. You'are right, the videos seem to be already cropped. I'm downloading the taichi dataset to do reconstruction for evaluation. Previously I followed the instruction in the picture to download the dataset. And it seems the downloaded dataset is already cropped to 256*256 ,right? And before I realized that, I followed this project page to pre-process the videos. By the way, if I want to train the model for taichi dataset, is additianal process still needed such as convert the format of .mp4 to that of .png?
Right the dataset is already cropped. Conversion is not strictly needed, and converting from mp4 to png won't make any sense because data is already lossy compressed. So if you want better performance, you should crop dataset in png format the first place.
I had similar problem even on original uncroped videos. Seems there is no implementation to save the cropped video or images. Adding the "save" function in util.py solves the problem.
Please correct me if I actually did something wrong.