Corpus ID: 220713216

Few-shot Visual Reasoning with Meta-analogical Contrastive Learning

  title={Few-shot Visual Reasoning with Meta-analogical Contrastive Learning},
  author={Youngsung Kim and Jinwoo Shin and Eunho Yang and Sung Ju Hwang},
While humans can solve a visual puzzle that requires logical reasoning by observing only few samples, it would require training over large amount of data for state-of-the-art deep reasoning models to obtain similar performance on the same task. In this work, we propose to solve such a few-shot (or low-shot) visual reasoning problem, by resorting to analogical reasoning, which is a unique human ability to identify structural or relational similarity between two sets. Specifically, given training… Expand

Figures and Tables from this paper

As a step towards improving the abstract reasoning capability of machines, we aim to solve Raven’s Progressive Matrices (RPM) with neural networks, since solving RPM puzzles is highly correlated withExpand
Contrastive Spatial Reasoning on Multi-View Line Drawings
This work uses a self-supervised binary classification network to compare the line drawing differences between various views of any two similar 3D objects and proposes a simple contrastive learning approach along with other network modifications to improve the baseline performance. Expand
Transductive Few-Shot Classification on the Oblique Manifold
This work proposes a non-parametric Region Self-attention with Spatial Pyramid Pooling (RSSPP), which realizes a trade-off between the generalization and the discriminative ability of the single image feature. Expand
A Broad Study on the Transferability of Visual Representations with Contrastive Learning
A comprehensive study on the transferability of learned representations of different contrastive approaches for linear evaluation, full-network transfer, and few-shot recognition on 12 downstream datasets from different domains, and object detection tasks on MSCOCO and VOC0712 shows that the contrasts learn representations that are easily transferable to a different downstream task. Expand
A Data Augmentation Method by Mixing Up Negative Candidate Answers for Solving Raven's Progressive Matrices
A data augmentation strategy by image mix-up is proposed, which is generalizable to a variety of multiple-choice problems, especially for image-based RPM-like problems and achieves significant and consistent improvement on various RPM- like datasets compared with the state-of-the-art models. Expand


RAVEN: A Dataset for Relational and Analogical Visual REasoNing
This work proposes a new dataset, built in the context of Raven's Progressive Matrices (RPM) and aimed at lifting machine intelligence by associating vision with structural, relational, and analogical reasoning in a hierarchical representation and establishes a semantic link between vision and reasoning by providing structure representation. Expand
Learning to Make Analogies by Contrasting Abstract Relational Structure
This work studies how analogical reasoning can be induced in neural networks that learn to perceive and reason about raw visual data and finds that the critical factor for inducing such a capacity is not an elaborate architecture, but rather, careful attention to the choice of data and the manner in which it is presented to the model. Expand
Abstract Reasoning with Distracting Features
This paper proposes feature robust abstract reasoning (FRAR) model, which consists of a reinforcement learning based teacher network to determine the sequence of training and a student network for predictions that is able to beat the state-of-the-art models. Expand
Deep Visual Analogy-Making
A novel deep network trained end-to-end to perform visual analogy making, which is the task of transforming a query image according to an example pair of related images, is developed. Expand
A simple neural network module for relational reasoning
This work shows how a deep learning architecture equipped with an RN module can implicitly discover and learn to reason about entities and their relations. Expand
Measuring abstract reasoning in neural networks
A dataset and challenge designed to probe abstract reasoning, inspired by a well-known human IQ test, is proposed and ways to both measure and induce stronger abstract reasoning in neural networks are introduced. Expand
Are Disentangled Representations Helpful for Abstract Visual Reasoning?
A large-scale study that investigates whether disentangled representations are more suitable for abstract reasoning tasks and observes that disentangle representations do in fact lead to better down-stream performance and enable quicker learning using fewer samples. Expand
Learning Perceptual Inference by Contrasting
It is demonstrated that CoPINet sets the new state-of-the-art for permutation-invariant models on two major datasets and concludes that spatial-temporal reasoning depends on envisaging the possibilities consistent with the relations between objects and can be solved from pixel-level inputs. Expand
Analogy-preserving Semantic Embedding for Visual Object Categorization
Analogy-preserving Semantic Embedding (ASE) is proposed to model analogies that reflect the relationships between multiple pairs of classes simultaneously, in the form "p is to q, as r is to s". Expand
Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
We propose an algorithm for meta-learning that is model-agnostic, in the sense that it is compatible with any model trained with gradient descent and applicable to a variety of different learningExpand