Hi ,
Thanks for sharing the awesome work! I've noticed the code seems to be designed for pre-processed data. Is there a way to use raw videos as input? If so, could you provide some guidance or point me to relevant resources?
Any help would be greatly appreciated!
Hi, raw videos were used only to get the hand poses. However, they can be used as a dataset for multiple downstream problems when combined with pose annotations.
Hi , Thanks for sharing the awesome work! I've noticed the code seems to be designed for pre-processed data. Is there a way to use raw videos as input? If so, could you provide some guidance or point me to relevant resources? Any help would be greatly appreciated!