A simple neural network module for relational reasoning

Code

• github.com/gngdb/relational
  torch★ 0
• github.com/rishikmani/sort-of-clevr
  tf★ 0
• github.com/kimhc6028/relational-networks
  pytorch★ 0
• github.com/yanchunyu71/relational-networks-paddle
  paddle★ 0
• github.com/AndreaCossu/Relation-Network-PyTorch
  pytorch★ 0
• github.com/clvrai/relation-network-tensorflow
  tf★ 0
• github.com/rosinality/relation-networks-pytorch
  pytorch★ 0
• github.com/moduIo/Relation-Networks
  tf★ 0
• github.com/ttok0s7u2n5/ML2_proj
  pytorch★ 0
• github.com/nerdimite/relation-network
  pytorch★ 0

Abstract

Relational reasoning is a central component of generally intelligent behavior, but has proven difficult for neural networks to learn. In this paper we describe how to use Relation Networks (RNs) as a simple plug-and-play module to solve problems that fundamentally hinge on relational reasoning. We tested RN-augmented networks on three tasks: visual question answering using a challenging dataset called CLEVR, on which we achieve state-of-the-art, super-human performance; text-based question answering using the bAbI suite of tasks; and complex reasoning about dynamic physical systems. Then, using a curated dataset called Sort-of-CLEVR we show that powerful convolutional networks do not have a general capacity to solve relational questions, but can gain this capacity when augmented with RNs. Our work shows how a deep learning architecture equipped with an RN module can implicitly discover and learn to reason about entities and their relations.

Tasks

Benchmark Results

Reproductions