Abstract: Humans are capable of inferring dynamic context from a still image and, with the provision of additional commonsense knowledge, can accurately complete visual commonsense reasoning tasks.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results