Systematic generalization is the ability to combine known parts into novel meaning; an important aspect of efficient human learning, but a weakness of neural network learning. In this work, we investigate how two well-known modeling principles -- modularity and data augmentation -- affect systematic generalization of neural networks in grounded language learning. We analyze how large the vocabulary needs to be to achieve systematic generalization and how similar the augmented data needs to be to the problem at hand. Our findings show that even in the controlled setting of a synthetic benchmark, achieving systematic generalization remains very difficult. After training on an augmented dataset with almost forty times more adverbs than the original problem, a non-modular baseline is not able to systematically generalize to a novel combination of a known verb and adverb. When separating the task into cognitive processes like perception and navigation, a modular neural network is able to utilize the augmented data and generalize more systematically, achieving 70% and 40% exact match increase over state-of-the-art on two gSCAN tests that have not previously been improved. We hope that this work gives insight into the drivers of systematic generalization, and what we still need to improve for neural networks to learn more like humans do.
GenKnowSub: Improving Modularity and Reusability of LLMs through General Knowledge Subtraction
School of Electrical and Computer Engineering, College of Engineering, University of Tehran, Tehran, Iran(伊朗德黑兰大学电气与计算机工程学院);Tehran Institute for Advanced Studies, Khatam University, Tehran, Iran(德黑兰高级研究学院)
UU-Tax at SemEval-2022 Task 3: Improving the generalizability of language models for taxonomy classification through data augmentation
Utrecht University(乌得勒支大学);Arab Academy for Science, Technology, and Maritime Transport(阿拉伯科技与海运学院);Leiden University Medical Center(莱顿大学医学中心);Leiden University(莱顿大学)
MoireMix: A Formula-Based Data Augmentation for Improving Image Classification Robustness
National Institute of Advanced Industrial Science and Technology (AIST)(国立研究开发法人产业技术综合研究所);Tokyo Denki University(东京电机大学);Waseda University(早稻田大学);Fundamental AI Lab, University of Technology Nuremberg(纽伦堡工业大学基础人工智能实验室);Visual Geometry Group, University of Oxford(牛津大学视觉几何组)
Improving age prediction: Utilizing LSTM-based dynamic forecasting for data augmentation in multivariate time series analysis
Tri-Institutional Center for Translational Research in Neuroimaging and Data Science (TReNDS): Georgia State University, Georgia Institute of Technology, Emory University(转化神经影像与数据科学三机构中心(TReNDS):佐治亚州立大学、佐治亚理工学院、埃默里大学)
Improving Transferability of Representations via Augmentation-Aware Self-Supervision
Korea Advanced Institute of Science and Technology (KAIST)(韩国科学技术院);University of Michigan(密歇根大学);University of California, Berkeley(加州大学伯克利分校);Amazon Web Services(亚马逊云科技);LG AI Research(LG人工智能研究院)
TTA-OOD: Test-time Augmentation for Improving Out-of-Distribution Detection in Gastrointestinal Vision
Nepal Applied Mathematics and Informatics Institute for research(NAAMII)(尼泊尔应用数学与信息学研究所(NAAMII));Fogsphere(Redev AI Ltd.)(Fogsphere(Redev AI有限公司));School of Natural and Computing Sciences, University of Aberdeen(阿伯丁大学自然与计算科学学院);West Virginia University(西弗吉尼亚大学)
Improving Behavioural Cloning with Human-Driven Dynamic Dataset Augmentation
School of Computing, University of Eastern Finland(东芬兰大学计算学院);Department of Electrical and Computer Engineering, National University of Singapore(新加坡国立大学电气与计算机工程系)
SA-Attack: Improving Adversarial Transferability of Vision-Language Pre-training Models via Self-Augmentation
Institute of Information Engineering, Chinese Academy of Sciences(中国科学院信息工程研究所);Nanyang Technological University(南洋理工大学);National University of Singapore(新加坡国立大学);Sun Yat-sen University(中山大学)
Improving Code-Switching Speech Recognition with TTS Data Augmentation
Institute for Infocomm Research (I 2 R), A*STAR, Singapore(信息通信研究所(I 2 R),A*STAR,新加坡);College of Computing and Data Science, Nanyang Technological University, Singapore(计算与数据科学学院,南洋理工大学,新加坡)
Improving End-to-End Speech Recognition for Dysarthric Speech through In-Domain Data Augmentation
Department of Electronics and Communication Engineering, National Institute of Technology Sikkim, India(电子与通信工程系,印度尼特拉特技术学院Sikkim分校);Signal Analysis and Interpretation Laboratory (SAIL), University of Southern California, Los Angeles, USA(信号分析与解释实验室(SAIL),美国南加州大学洛杉矶分校)
Improving End-To-End Modeling for Mispronunciation Detection with Effective Augmentation Mechanisms
National Taiwan Normal University(台湾师范大学)
近年来,端到端(E2 E)模型,它允许采取的频谱向量序列的第二语言(第二语言)学习者的话语作为输入,并产生相应的音素级序列作为输出,引起了研究的关注,在开发的发音错误检测(MD)系统。然而,由于缺乏足够的L2扬声器的标记语音数据用于模型估计,E2 E MD模型相对于建立在DNN-HMM声学模型上的传统模型容易过拟合。为了缓解这一关键问题,我们在本文中提出了两种建模策略,以提高E2 E MD模型的区分能力,每一种都可以隐式地利用编码在预训练声学模型中的语音和音韵特征,并分别包含在训练数据的参考转录本中。第一个是输入增强,其目的是从DNN-HMM声学模型中提取有关语音识别的知识。第二种是标签增强,它设法从训练数据的转录本中捕获更多的语音模式。在L2-ARCTIC英语数据集上进行的一系列实证实验似乎证实了我们的E2 E MD模型与一些顶级E2 E MD模型和基于DNN-HMM声学模型的经典发音评分方法相比的有效性。
Improving Outdoor Multi-cell Fingerprinting-based Positioning via Mobile Data Augmentation
CNIT/WiLab - National Wireless Communication Laboratory(CNIT/WiLab-国家无线通信实验室);Department of Information Engineering, Electronics and Telecommunications, Sapienza University of Rome(信息工程、电子与电信系,萨皮恩扎罗马大学)
Improving Quality Control of Whole Slide Images by Explicit Artifact Augmentation
AGH University of Krakow(克拉科夫AGH科技大学);University of Applied Sciences Western Switzerland (HES-SO)(瑞士西部应用科学大学(HES-SO));Radboud University Medical Center(拉德堡德大学医学中心);University of Padova(帕多瓦大学);University of Geneva(日内瓦大学)