{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,7,30]],"date-time":"2025-07-30T16:05:33Z","timestamp":1753891533656,"version":"3.41.2"},"reference-count":62,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T00:00:00Z","timestamp":1711411200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neurorobot."],"abstract":"<jats:p>Data augmentation is an effective technique for automatically expanding training data in deep learning. Brain-inspired methods are approaches that draw inspiration from the functionality and structure of the human brain and apply these mechanisms and principles to artificial intelligence and computer science. When there is a large style difference between training data and testing data, common data augmentation methods cannot effectively enhance the generalization performance of the deep model. To solve this problem, we improve modeling Domain Shifts with Uncertainty (DSU) and propose a new brain-inspired computer vision image data augmentation method which consists of two key components, namely, <jats:italic>using Robust statistics and controlling the Coefficient of variance for DSU<\/jats:italic> (RCDSU) and <jats:italic>Feature Data Augmentation<\/jats:italic> (FeatureDA). RCDSU calculates feature statistics (mean and standard deviation) with robust statistics to weaken the influence of outliers, making the statistics close to the real values and improving the robustness of deep learning models. By controlling the coefficient of variance, RCDSU makes the feature statistics shift with semantic preservation and increases shift range. FeatureDA controls the coefficient of variance similarly to generate the augmented features with semantics unchanged and increase the coverage of augmented features. RCDSU and FeatureDA are proposed to perform style transfer and content transfer in the feature space, and improve the generalization ability of the model at the style and content level respectively. On Photo, Art Painting, Cartoon, and Sketch (PACS) multi-style classification task, RCDSU plus FeatureDA achieves competitive accuracy. After adding Gaussian noise to PACS dataset, RCDSU plus FeatureDA shows strong robustness against outliers. FeatureDA achieves excellent results on CIFAR-100 image classification task. RCDSU plus FeatureDA can be applied as a novel brain-inspired semantic data augmentation method with implicit robot automation which is suitable for datasets with large style differences between training and testing data.<\/jats:p>","DOI":"10.3389\/fnbot.2024.1382406","type":"journal-article","created":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T04:22:09Z","timestamp":1711426929000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Brain-inspired semantic data augmentation for multi-style images"],"prefix":"10.3389","volume":"18","author":[{"given":"Wei","family":"Wang","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhaowei","family":"Shang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chengxing","family":"Li","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2024,3,26]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"1239581","DOI":"10.3389\/fnbot.2023.1239581","article-title":"Neurorobotic reinforcement learning for domains with parametrical uncertainty","volume":"17","author":"Amaya","year":"2023","journal-title":"Front. Neurorobot"},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1711.04340","article-title":"Data augmentation generative adversarial networks","author":"Antoniou","year":"2017","journal-title":"arXiv [Preprint]"},{"key":"B3","first-page":"169","article-title":"\u201cComputationally efficient robust sparse estimation in high dimensions,\u201d","volume-title":"Conference on Learning Theory","author":"Balakrishnan","year":"2017"},{"key":"B4","first-page":"3722","article-title":"\u201cUnsupervised pixel-level domain adaptation with generative adversarial networks,\u201d","author":"Bousmalis","year":"2016","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B5","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1810.10863","article-title":"Gan augmentation: augmenting training data using generative adversarial networks","author":"Bowles","year":"2018","journal-title":"arXiv [Preprint]."},{"key":"B6","doi-asserted-by":"publisher","first-page":"93","DOI":"10.3390\/fi14030093","article-title":"Survey on videos data augmentation for deep learning models","volume":"14","author":"Cauli","year":"2022","journal-title":"Future Internet"},{"key":"B7","first-page":"22405","article-title":"\u201cSwad: Domain Generalization by Seeking Flat Minima,\u201d","author":"Cha","year":"2021","journal-title":"Advances in Neural Information Processing Systems, Vol. 31"},{"key":"B8","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR42600.2020.00575","article-title":"\u201cData uncertainty learning in face recognition,\u201d","volume-title":"2020 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","author":"Chang","year":"2020"},{"key":"B9","first-page":"7318","article-title":"\u201cOutlier-robust Sparse Estimation via Non-convex Optimization,\u201d","author":"Cheng","year":"2022","journal-title":"Advances in Neural Information Processing Systems, Vol. 35"},{"key":"B10","first-page":"727","article-title":"\u201cFaster algorithms for high-dimensional robust covariance estimation,\u201d","volume-title":"Conference on Learning Theory","author":"Cheng","year":"2019"},{"key":"B11","article-title":"\u201cRobust learning of fixed-structure Bayesian networks,\u201d","volume-title":"Advances in Neural Information Processing Systems, Vol. 31","author":"Cheng","year":"2018"},{"key":"B12","doi-asserted-by":"publisher","first-page":"3537","DOI":"10.3390\/math11163537","article-title":"Promatch: semi-supervised learning with prototype consistency","volume":"11","author":"Cheng","year":"2023","journal-title":"Mathematics"},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1805.09501","article-title":"Autoaugment: learning augmentation policies from data","author":"Cubuk","year":"2018","journal-title":"arXiv [Preprint]. arXiv:1805.09501"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2304.13976","article-title":"Moderately distributional exploration for domain generalization","author":"Dai","year":"","journal-title":"arXiv [Preprint]"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2302.13007","article-title":"Auggpt: leveraging chatgpt for text data augmentation","author":"Dai","year":"","journal-title":"arXiv [Preprint]."},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.","article-title":"Improved regularization of convolutional neural networks with cutout","author":"DeVries","year":"2017","journal-title":"arXiv [Preprint]"},{"key":"B17","doi-asserted-by":"publisher","first-page":"568359","DOI":"10.3389\/fnbot.2020.568359","article-title":"Nengo and low-power ai hardware for robust, embedded neurorobotics","volume":"14","author":"DeWolf","year":"2020","journal-title":"Front. Neurorobot"},{"key":"B18","doi-asserted-by":"publisher","first-page":"742","DOI":"10.1137\/17M1126680","article-title":"Robust estimators in high-dimensions without the computational intractability","volume":"48","author":"Diakonikolas","year":"","journal-title":"SIAM J. Comput"},{"key":"B19","first-page":"1645","article-title":"\u201cOutlier-robust learning of Ising models under Dobrushin's condition,\u201d","volume-title":"Conference on Learning Theory","author":"Diakonikolas","year":"2021"},{"key":"B20","first-page":"2745","article-title":"\u201cEfficient algorithms and lower bounds for robust linear regression,\u201d","volume-title":"Proceedings of the Thirtieth Annual ACM-SIAM Symposium on Discrete Algorithms","author":"Diakonikolas","year":""},{"key":"B21","first-page":"1596","article-title":"\u201cSever: a robust meta-algorithm for stochastic optimization,\u201d","volume-title":"International Conference on Machine Learning","author":"Diakonikolas","year":""},{"key":"B22","article-title":"\u201cOutlier-robust High-dimensional Sparse Estimation via Iterative Filtering,\u201d","author":"Diakonikolas","year":"","journal-title":"Advances in Neural Information Processing Systems, Vol. 32"},{"key":"B23","doi-asserted-by":"crossref","first-page":"94","DOI":"10.1007\/978-3-658-29267-6_21","article-title":"\u201cDeep learning-based denoising of mammographic images using physics-driven data augmentation,\u201d","volume-title":"Bildverarbeitung f\u00fcr die Medizin 2020: Algorithmen-Systeme-Anwendungen. Proceedings des Workshops vom 15. bis 17. M\u00e4rz 2020 in Berlin","author":"Eckert","year":"2020"},{"key":"B24","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2212.10558","article-title":"On-the-fly denoising for data augmentation in natural language understanding","author":"Fang","year":"2022","journal-title":"arXiv [Preprint]"},{"key":"B25","doi-asserted-by":"publisher","first-page":"856797","DOI":"10.3389\/fnbot.2022.856797","article-title":"Evaluating muscle synergies with EMG data and physics simulation in the neurorobotics platform","volume":"16","author":"Feldotto","year":"2022","journal-title":"Front. Neurorobot"},{"key":"B26","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1506.02158","article-title":"Bayesian convolutional neural networks with Bernoulli approximate variational inference","author":"Gal","year":"2015","journal-title":"arXiv [Preprint]"},{"key":"B27","first-page":"1050","article-title":"\u201cDropout as a Bayesian approximation: representing model uncertainty in deep learning,\u201d","volume-title":"International Conference on Machine Learning","author":"Gal","year":"2016"},{"key":"B28","first-page":"104","article-title":"\u201cExtending temporal data augmentation for video action recognition,\u201d","volume-title":"International Conference on Image and Vision Computing New Zealand","author":"Gorpincenko","year":"2022"},{"key":"B29","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.90","article-title":"\u201cDeep residual learning for image recognition,\u201d","volume-title":"2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)","author":"He","year":"2016"},{"key":"B30","doi-asserted-by":"publisher","first-page":"8704","DOI":"10.1109\/TPAMI.2019.2918284","article-title":"Convolutional networks with dense connectivity","volume":"44","author":"Huang","year":"2019","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"B31","doi-asserted-by":"crossref","first-page":"124","DOI":"10.1007\/978-3-030-58536-5_8","article-title":"\u201cSelf-challenging improves cross-domain generalization,\u201d","volume-title":"Computer Vision ECCV 2020","author":"Huang","year":"2020"},{"key":"B32","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s11263-015-0823-z","article-title":"Reading text in the wild with convolutional neural networks","volume":"116","author":"Jaderberg","year":"2015","journal-title":"Int. J. Comput. Vis"},{"key":"B33","doi-asserted-by":"crossref","first-page":"80","DOI":"10.1109\/VIS54862.2022.00025","article-title":"\u201cUniform manifold approximation with two-phase optimization,\u201d","volume-title":"2022 IEEE Visualization and Visual Analytics (VIS)","author":"Jeon","year":"2022"},{"key":"B34","article-title":"\u201cWhat uncertainties do we need in Bayesian deep learning for computer vision?\u201d","volume-title":"Advances in Neural Information Processing Systems, Vol. 30","author":"Kendall","year":"2017"},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2206.15015","article-title":"Exploring temporally dynamic data augmentation for video recognition","author":"Kim","year":"2022","journal-title":"arXiv [Preprint]."},{"key":"B36","first-page":"1420","article-title":"\u201cEfficient algorithms for outlier-robust regression,\u201d","volume-title":"Conference On Learning Theory","author":"Klivans","year":"2018"},{"key":"B37","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1145\/3065386","article-title":"Imagenet classification with deep convolutional neural networks","volume":"60","author":"Krizhevsky","year":"2017","journal-title":"Commun. ACM"},{"journal-title":"Learning Multiple Layers of Features from Tiny Images","year":"2009","author":"Krizhevsky","key":"B38"},{"key":"B39","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2202.03958","article-title":"Uncertainty modeling for out-of-distribution generalization","author":"Li","year":"2022","journal-title":"arXiv [Preprint]."},{"key":"B40","article-title":"\u201cLearning raw image denoising with bayer pattern unification and bayer preserving augmentation,\u201d","volume-title":"2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)","author":"Liu","year":"2020"},{"key":"B41","doi-asserted-by":"publisher","first-page":"1753","DOI":"10.32604\/iasc.2023.039600","article-title":"Attentive neighborhood feature augmentation for semi-supervised learning","volume":"37","author":"Liu","year":"2023","journal-title":"Intell. Autom. Soft Comput"},{"key":"B42","doi-asserted-by":"publisher","first-page":"1269","DOI":"10.3390\/electronics10111269","article-title":"GPR B-scan image denoising via multi-scale convolutional autoencoder with data augmentation","volume":"10","author":"Luo","year":"2021","journal-title":"Electronics"},{"volume-title":"Robust Statistics: Theory and Methods (with R)","year":"2019","author":"Maronna","key":"B43"},{"key":"B44","first-page":"9482","article-title":"\u201cPermuted ADaIN: reducing the bias towards global statistics in image classification,\u201d","author":"Nuriel","year":"2021","journal-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition"},{"key":"B45","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2009.12976","article-title":"Robust regression with covariate filtering: heavy tails and adversarial contamination","author":"Pensia","year":"2020","journal-title":"arXiv [Preprint]."},{"key":"B46","doi-asserted-by":"publisher","first-page":"601","DOI":"10.1111\/rssb.12364","article-title":"Robust estimation via robust gradient estimation","volume":"82","author":"Prasad","year":"2020","journal-title":"J. R. Stat. Soc. B: Stat. Methodol"},{"key":"B47","doi-asserted-by":"publisher","first-page":"1243","DOI":"10.32604\/iasc.2023.038241","article-title":"Instance reweighting adversarial training based on confused label","volume":"37","author":"Qiu","year":"2023","journal-title":"Intell. Autom. Soft Comput"},{"key":"B48","article-title":"\u201cLearning to compose domain-specific transformations for data augmentation,\u201d","volume-title":"Advances in Neural Information Processing Systems, Vol. 30","author":"Ratner","year":"2017"},{"key":"B49","doi-asserted-by":"publisher","first-page":"73","DOI":"10.1002\/widm.2","article-title":"Robust statistics for outlier detection","volume":"1","author":"Rousseeuw","year":"2011","journal-title":"Wiley Interdiscip. Rev.: Data Min. Knowl. Discov"},{"key":"B50","article-title":"\u201cProbabilistic face embeddings,\u201d","volume-title":"2019 IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Shi","year":"2020"},{"key":"B51","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1409.1556","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2014","journal-title":"arXiv [Preprint]"},{"key":"B52","article-title":"\u201cTraining Very Deep Networks,\u201d","author":"Srivastava","year":"2015","journal-title":"Advances in Neural Information Processing Systems, Vol. 28"},{"key":"B53","doi-asserted-by":"publisher","first-page":"12","DOI":"10.1016\/j.jpdc.2019.03.003","article-title":"The security of machine learning in an adversarial setting: a survey","volume":"130","author":"Wang","year":"2019","journal-title":"J. Parallel Distributed Comput"},{"key":"B54","doi-asserted-by":"publisher","first-page":"3733","DOI":"10.1109\/TPAMI.2021.3052951","article-title":"Regularizing deep networks with semantic data augmentation","volume":"44","author":"Wang","year":"2021","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"B55","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/D19-1670","article-title":"\u201cEDA: easy data augmentation techniques for boosting performance on text classification tasks,\u201d","volume-title":"Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)","author":"Wei","year":"2019"},{"key":"B56","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2202.13840","article-title":"Text smoothing: enhance various data augmentation methods on text classification tasks","author":"Wu","year":"2022","journal-title":"arXiv"},{"key":"B57","first-page":"7097","article-title":"\u201cPCL: proxy-based contrastive learning for domain generalization,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Yao","year":"2022"},{"key":"B58","article-title":"\u201cRobust person re-identification by modelling feature uncertainty,\u201d","volume-title":"2019 IEEE\/CVF International Conference on Computer Vision (ICCV)","author":"Yu","year":"2020"},{"key":"B59","doi-asserted-by":"crossref","first-page":"034002","DOI":"10.1088\/2634-4386\/ace64c","article-title":"Brain-inspired methods for achieving robust computation in heterogeneous mixed-signal neuromorphic processing systems","volume":"3","author":"Zendrikov","year":"2023","journal-title":"Neuromorphic Comput. Eng"},{"key":"B60","doi-asserted-by":"publisher","first-page":"13001","DOI":"10.1609\/aaai.v34i07.7000","article-title":"Random erasing data augmentation","volume":"34","author":"Zhong","year":"2020","journal-title":"Proc. AAAI Conf. Artif. Intell"},{"key":"B61","doi-asserted-by":"crossref","first-page":"561","DOI":"10.1007\/978-3-030-58517-4_33","article-title":"\u201cLearning to generate novel domains for domain generalization,\u201d","volume-title":"Computer Vision ECCV 2020: 16th European Conference, Glasgow, UK, August 23\u201328, 2020, Proceedings, Part XVI","author":"Zhou","year":"2020"},{"key":"B62","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2104.02008","article-title":"Domain generalization with mixstyle","author":"Zhou","year":"2021","journal-title":"arXiv [Preprint]"}],"container-title":["Frontiers in Neurorobotics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2024.1382406\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,3,26]],"date-time":"2024-03-26T04:22:38Z","timestamp":1711426958000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2024.1382406\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,3,26]]},"references-count":62,"alternative-id":["10.3389\/fnbot.2024.1382406"],"URL":"https:\/\/doi.org\/10.3389\/fnbot.2024.1382406","relation":{},"ISSN":["1662-5218"],"issn-type":[{"type":"electronic","value":"1662-5218"}],"subject":[],"published":{"date-parts":[[2024,3,26]]},"article-number":"1382406"}}