The role of object-centric representations, guided attention, and external memory on generalizing visual relations. (arXiv:2304.07091v1 [cs.CV])

Visual reasoning is a long-term goal of vision research. In the last decade,
several works have attempted to apply deep neural networks (DNNs) to the task
of learning visual relations from images, with modest results in terms of the
generalization of the relations learned. In recent years, several innovations
in DNNs have been developed in order to enable learning abstract relation from
images. In this work, we systematically evaluate a series of DNNs that
integrate mechanism such as slot attention, …

apply arxiv attention images innovations long-term memory networks neural networks reasoning relations research role series terms vision vision research work