With no explanation, label text_A→text_B with either "strengthener" or "weakener".
text_A: A child carrying a large object on his head while looking at the camera. The child is carrying a water jug on his head.
text_B: The object is in the shape of a square.
weakener.