DuyNguDao
commented
2 months ago Hi Zubair We create the datasets used for action recognition through real cameras and a few datasets on the internet. We created a few tools with Python to process video data as extract action, ..vv. We filtered this data and public on the Kaggle web and referred to the source video then we are used. We lose three months to collect, record action videos, and process it. You can write a tool to processing videos with python (ex. opencv lib). Thank you for your feedback. Best regards, Dao Duy Ngu
Vào Th 7, 17 thg 8, 2024 lúc 21:13 Zubair Khan @.***> đã viết:
Thankyou for excellent sharing. Can you please tell the exact datasets used for action recognition. i.e. URFall and others. Also, it is necessary to trim the videos into clips to only have the specific action and not the other activity. right?
— Reply to this email directly, view it on GitHub https://github.com/DuyNguDao/Identity-Action/issues/7, or unsubscribe https://github.com/notifications/unsubscribe-auth/AUZ2UERL76YYTTBFWW4BX63ZR5LCLAVCNFSM6AAAAABMVOSIAGVHI2DSMVQWIX3LMV43ASLTON2WKOZSGQ3TCNJUGEZDENA . You are receiving this because you are subscribed to this thread.Message ID: @.***>
Thankyou for excellent sharing. Can you please tell the exact datasets used for action recognition. i.e. URFall and others. Also, it is necessary to trim the videos into clips to only have the specific action and not the other activity. right?