Unsupervised Abstract Reasoning for Raven’s Problem Matrices
Unsupervised Abstract Reasoning for Raven’s Problem Matrices
复制标题
DOI：
10.1109/tip.2021.3114987
复制
发表时间：
2021-09
影响因子：
10.6
通讯作者：
Tao Zhuo;Qian Huang;Mohan S. Kankanhalli
中科院分区：
文献类型：
--
作者：
Tao Zhuo;Qian Huang;Mohan S. Kankanhalli
瑞文推理测验（RPM）与人类智力高度相关，被广泛用于测量人类的抽象推理能力。在本文中，为了研究深度神经网络的抽象推理能力，我们提出了第一种用于解决RPM问题的无监督学习方法。由于不允许使用地面真值标签，因此我们根据RPM公式的先验约束设计了一个伪目标来近似地面真值标签，从而有效地将无监督学习策略转换为有监督学习策略。然而，正确的答案是错误的标记的伪目标，从而噪声对比将导致不准确的模型训练。为了缓解这个问题，我们建议提高模型的性能与否定的答案。此外，我们开发了一个分散的方法，以适应不同的RPM问题的特征表示。在三个数据集上的大量实验表明，我们的方法甚至优于一些监督方法。我们的代码可在https://github.com/visiontao/ncd上获得。
Raven’s Progressive Matrices (RPM) is highly correlated with human intelligence, and it has been widely used to measure the abstract reasoning ability of humans. In this paper, to study the abstract reasoning capability of deep neural networks, we propose the first unsupervised learning method for solving RPM problems. Since the ground truth labels are not allowed, we design a pseudo target based on the prior constraints of the RPM formulation to approximate the ground-truth label, which effectively converts the unsupervised learning strategy into a supervised one. However, the correct answer is wrongly labelled by the pseudo target, and thus the noisy contrast will lead to inaccurate model training. To alleviate this issue, we propose to improve the model performance with negative answers. Moreover, we develop a decentralization method to adapt the feature representation to different RPM problems. Extensive experiments on three datasets demonstrate that our method even outperforms some of the supervised approaches. Our code is available at https://github.com/visiontao/ncd.