I'm wondering how the annotations were done in VRD dataset.
The "category": 0 for subject or object means that the entity is a person?
And the sequence just goes down as the Table 4 in the paper?
So could I infer that "category": 1 for truck and "category": 2 for shirt ?
Is it the same implication that the Table 3 for the predicate goes in the same way?
I'm wondering how the annotations were done in VRD dataset.
The "category": 0 for subject or object means that the entity is a person? And the sequence just goes down as the Table 4 in the paper? So could I infer that "category": 1 for truck and "category": 2 for shirt ?
Is it the same implication that the Table 3 for the predicate goes in the same way?