I have a set of videos which range between 3 to 10 minutes. In these videos, multiple actions occur. An action is roughly around 2 seconds.
I need to know the exact timestamps in the videos of where the actions occur (start and stop time), such that I can automate the clipping process afterwards.
Does anyone have insights on how to design the layout of this task? I will need text boxes where the time stamps can be entered, but the amount of actions that occur in the videos differ.
I personally thought about a video classification task, since that seems the only way to handle video input? But then the crowd-text-area seems to be limited in its capabilities, there I can only create columns (and not add different rows to each column).
Any insights are welcome
I need to know the exact timestamps in the videos of where the actions occur (start and stop time), such that I can automate the clipping process afterwards.
Does anyone have insights on how to design the layout of this task? I will need text boxes where the time stamps can be entered, but the amount of actions that occur in the videos differ.
I personally thought about a video classification task, since that seems the only way to handle video input? But then the crowd-text-area seems to be limited in its capabilities, there I can only create columns (and not add different rows to each column).
Any insights are welcome