Thank you Dr. Alexandros, I have 2 question how many activities were there from those three datasets that the experiment tried to identify? Can we predict multiple people's action in the same video frame? and Can we predict the path of an action before they make move: like a person taking out a chocolate bar from pocket, what his path would be after that or may be prediction of those path data from the video frame?

Can we use CO3D for high speed moving smaller object classification from video? Can we use it for 3D object tracking?

Yes, thank you Dr. Alexandros

Thank you for the presentation Dr. Alexandros

can we use same architecture for understanding activity and background scene together? If so how can we do it both things using same dataset?

Thank You Dr. Iosifidis

context

Is there any effect of noise on the action recognition? How much noise is tolerable for effective detection?

Yes Dr. Alexandros noise in video frame

thank you Dr. Losifidis

Thank you Dr. Alexandros