{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T18:26:31Z","timestamp":1787336791977,"version":"build-2736575974"},"reference-count":45,"publisher":"MIT Press","issue":"9","content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2026,8,21]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>A key challenge in visual object recognition is developing models that generalize from limited data while maintaining transparency in their decision making. We propose a biologically inspired model that addresses both issues by classifying images based on transformation-invariant local shape key features. Following the principles of the brain\u2019s what and where pathways, each feature is encoded by an image patch and its relative location in polar coordinates, enabling interpretable and robust comparisons between inputs and class prototypes. To mimic human concept learning, prototypes are selected using clustering, improving representativeness and generalization. Results show that our model achieves human-comparable performance, with an error rate between 1% and 2% on the MNIST data set when all training images are used as prototypes. In data-limited scenarios, where only a small number of prototypes are selected, our model consistently outperforms convolutional neural networks (CNNs). To evaluate out-of-distribution generalization, we use prototypes from MNIST and test both models on the ETL-1 data set, which differs in data distribution. Although CNN accuracy drops significantly under these conditions, our model maintains high accuracy, even with few prototypes, demonstrating strong robustness and greater capacity to generalize to unseen distributions, bringing it closer to human-like recognition capabilities.<\/jats:p>","DOI":"10.1162\/neco.a.1563","type":"journal-article","created":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:35:29Z","timestamp":1785537329000},"page":"1611-1632","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":0,"title":["Shape Matters: Few-Shot Object Classification From High-Information Contour Features"],"prefix":"10.1162","volume":"38","author":[{"given":"Maria","family":"Os\u00f3rio","sequence":"first","affiliation":[{"name":"INESC-ID, Instituto Superior T\u00e9cnico, Universidade de Lisboa, Lisbon 2744-016, Portugal maria.osorio@tecnico.ulisboa.pt"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexandre","family":"Bernardino","sequence":"additional","affiliation":[{"name":"Institute for Systems and Robotics, Instituto Superior T\u00e9cnico, Universidade de Lisboa, Lisbon 2744-016, Portugal alex@isr.tecnico.ulisboa.pt"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andreas","family":"Wichert","sequence":"additional","affiliation":[{"name":"INESC-ID, Instituto Superior T\u00e9cnico, Universidade de Lisboa, Lisbon 2744-016, Portugal andreas.wichert@tecnico.ulisboa.pt"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"281","published-online":{"date-parts":[[2026,8,21]]},"reference":[{"key":"2026082113304844600_bib1","doi-asserted-by":"publisher","first-page":"102475","DOI":"10.1109\/ACCESS.2022.3208603","article-title":"Representation learning in sensory cortex: A theory","volume":"10","author":"Anselmi","year":"2022","journal-title":"IEEE Access"},{"issue":"3","key":"2026082113304844600_bib2","doi-asserted-by":"publisher","first-page":"183","DOI":"10.1037\/h0054663","article-title":"Some informational aspects of visual perception","volume":"61","author":"Attneave","year":"1954","journal-title":"Psychological Review"},{"key":"2026082113304844600_bib3","article-title":"Combining different V1 brain model variants to improve robustness to image corruptions in CNNs","author":"Baidya","year":"2021","journal-title":"CoRR"},{"issue":"2","key":"2026082113304844600_bib4","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1037\/0033-295X.94.2.115","article-title":"Recognition-by-components: A theory of human image understanding","volume":"94","author":"Biederman","year":"1987","journal-title":"Psychological Review"},{"key":"2026082113304844600_bib5","first-page":"1597","article-title":"A simple framework for contrastive learning of visual representations","volume-title":"Proceedings of the 37th International Conference on Machine Learning","author":"Chen","year":"2020"},{"key":"2026082113304844600_bib6","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN.1990.137720","author":"Darken","year":"1990","journal-title":"Fast adaptive k-means clustering: Some empirical results"},{"key":"2026082113304844600_bib7","article-title":"ETL1","author":"ETL Character Database","year":"2025"},{"issue":"1","key":"2026082113304844600_bib8","doi-asserted-by":"publisher","first-page":"243","DOI":"10.1037\/0033-295X.112.1.243","article-title":"Information along contours and object boundaries","volume":"112","author":"Feldman","year":"2005","journal-title":"Psychological Review"},{"key":"2026082113304844600_bib9","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2010.5539906","article-title":"Cascade object detection with deformable part models","volume-title":"Proceedings of the 2010 Computer Society Conference on Computer Vision and Pattern Recognition","author":"Felzenszwalb","year":"2010"},{"issue":"4","key":"2026082113304844600_bib10","doi-asserted-by":"publisher","first-page":"193","DOI":"10.1007\/BF00344251","article-title":"Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position","volume":"36","author":"Fukushima","year":"1980","journal-title":"Biological Cybernetics"},{"key":"2026082113304844600_bib11","article-title":"ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness","author":"Geirhos","year":"2018","journal-title":"Proceedings of the International Conference on Learning Representations"},{"key":"2026082113304844600_bib12","volume-title":"Deep learning","author":"Goodfellow","year":"2016"},{"issue":"1537","key":"2026082113304844600_bib13","doi-asserted-by":"publisher","first-page":"177","DOI":"10.1098\/rstb.2009.0200","article-title":"Learning to represent visual input","volume":"365","author":"Hinton","year":"2010","journal-title":"Philosophical Transactions of the Royal Society B: Biological Sciences"},{"key":"2026082113304844600_bib14","volume-title":"Eye, brain, and vision","author":"Hubel","year":"1995"},{"issue":"1","key":"2026082113304844600_bib15","doi-asserted-by":"publisher","first-page":"106","DOI":"10.1113\/jphysiol.1962.sp006837","article-title":"Receptive fields, binocular interaction and functional architecture in the cat\u2019s visual cortex","volume":"160","author":"Hubel","year":"1962","journal-title":"Journal of Physiology"},{"key":"2026082113304844600_bib16","first-page":"69","article-title":"Lois de distribution florale dans la zone alpine","volume":"38","author":"Jaccard","year":"1902","journal-title":"Bulletin de la Soci\u00e9t\u00e9 vaudoise des sciences naturelles"},{"issue":"4","key":"2026082113304844600_bib17","doi-asserted-by":"publisher","first-page":"490","DOI":"10.1037\/0096-1523.11.4.490","article-title":"Reading rotated words","volume":"11","author":"Koriat","year":"1985","journal-title":"Journal of Experimental Psychology: Human Perception and Performance"},{"key":"2026082113304844600_bib18","first-page":"1106","article-title":"ImageNet classification with deep convolutional neural networks","volume-title":"Advances in neural information processing systems","author":"Krizhevsky","year":"2012"},{"issue":"48","key":"2026082113304844600_bib19","doi-asserted-by":"publisher","first-page":"13232","DOI":"10.1523\/JNEUROSCI.3481-07.2007","article-title":"Magnocellular projections as the trigger of top-down facilitation in recognition","volume":"27","author":"Kveraga","year":"2007","journal-title":"Journal of Neuroscience"},{"issue":"3","key":"2026082113304844600_bib20","doi-asserted-by":"publisher","first-page":"299","DOI":"10.1016\/0885-2014(88)90014-7","article-title":"The importance of shape in early lexical learning","volume":"3","author":"Landau","year":"1988","journal-title":"Cognitive Development"},{"key":"2026082113304844600_bib21","author":"LeCun","year":"2010","journal-title":"MNIST handwritten digit database."},{"issue":"5","key":"2026082113304844600_bib22","doi-asserted-by":"publisher","first-page":"401","DOI":"10.1016\/S0960-9822(00)00089-0","article-title":"View-dependent object recognition by monkeys","volume":"4","author":"Logothetis","year":"1994","journal-title":"Current Biology"},{"issue":"2","key":"2026082113304844600_bib23","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1023\/B:VISI.0000029664.99615.94","article-title":"Distinctive image features from scale-invariant keypoints","volume":"60","author":"Lowe","year":"2004","journal-title":"International Journal of Computer Vision"},{"key":"2026082113304844600_bib24","doi-asserted-by":"crossref","first-page":"90","DOI":"10.1109\/GMAG.2003.1219671","article-title":"Geometric design and space planning using the marching squares and marching cube algorithms","volume-title":"Proceedings of the 2003 International Conference on Geometric Modeling and Graphic","author":"Maple","year":"2003"},{"key":"2026082113304844600_bib25","doi-asserted-by":"crossref","DOI":"10.7551\/mitpress\/9780262514620.001.0001","volume-title":"Vision: A computational investigation into the human representation and processing of visual information","author":"Marr","year":"2010"},{"key":"2026082113304844600_bib26","doi-asserted-by":"publisher","first-page":"414","DOI":"10.1016\/0166-2236(83)90190-X","article-title":"Object vision and spatial vision: Two cortical pathways","volume":"6","author":"Mishkin","year":"1983","journal-title":"Trends in Neurosciences"},{"issue":"6583","key":"2026082113304844600_bib27","doi-asserted-by":"publisher","first-page":"607","DOI":"10.1038\/381607a0","article-title":"Emergence of simple-cell receptive field properties by learning a sparse code for natural images","volume":"381","author":"Olshausen","year":"1996","journal-title":"Nature"},{"issue":"8","key":"2026082113304844600_bib28","doi-asserted-by":"crossref","first-page":"1626","DOI":"10.1162\/neco_a_01677","article-title":"Promoting the shift from pixel-level correlations to object semantics learning by rethinking computer vision benchmark data sets","volume":"36","author":"Os\u00f3rio","year":"2024","journal-title":"Neural Computation"},{"issue":"285\u2013296","key":"2026082113304844600_bib29","first-page":"23","article-title":"A threshold selection method from gray-level histograms","volume":"11","author":"Otsu","year":"1975","journal-title":"Automatica"},{"key":"2026082113304844600_bib30","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2506.03089","article-title":"Explicitly modeling subcortical vision with a neuro-inspired front-end improves CNN robustness","author":"Piper","year":"2025","journal-title":"CoRR"},{"issue":"3","key":"2026082113304844600_bib31","doi-asserted-by":"publisher","first-page":"244","DOI":"10.1016\/S0146-664X(72)80017-0","article-title":"An iterative procedure for the polygonal approximation of plane curves","volume":"1","author":"Ramer","year":"1972","journal-title":"Computer Graphics and Image Processing"},{"key":"2026082113304844600_bib32","volume-title":"The illusion of reality","author":"Resnikoff","year":"2012"},{"key":"2026082113304844600_bib33","doi-asserted-by":"publisher","first-page":"1019","DOI":"10.1038\/14819","article-title":"Hierarchical models of object recognition in cortex","volume":"2","author":"Riesenhuber","year":"1999","journal-title":"Nature Neuroscience"},{"issue":"4","key":"2026082113304844600_bib34","doi-asserted-by":"publisher","first-page":"719","DOI":"10.1037\/0096-1523.7.4.719","article-title":"The effect on form perception of change of orientation in the third dimension","volume":"7","author":"Rock","year":"1981","journal-title":"Journal of Experimental Psychology: Human Perception and Performance"},{"key":"2026082113304844600_bib35","doi-asserted-by":"publisher","first-page":"38","DOI":"10.1016\/J.NEUNET.2019.01.018","article-title":"Attention inspired network: Steep learning curve in an invariant pattern recognition model","volume":"114","author":"Sa-Couto","year":"2019","journal-title":"Neural Networks"},{"issue":"12","key":"2026082113304844600_bib36","doi-asserted-by":"publisher","first-page":"3334","DOI":"10.1162\/neco_a_01446","article-title":"Simple convolutional-based models: Are they learning the task or the data?","volume":"33","author":"Sa-Couto","year":"2021","journal-title":"Neural Computation"},{"key":"2026082113304844600_bib37","doi-asserted-by":"publisher","first-page":"97","DOI":"10.1016\/j.neucom.2022.04.130","article-title":"Using brain inspired principles to unsupervisedly learn good representations for visual pattern recognition","volume":"495","author":"Sa-Couto","year":"2022","journal-title":"Neurocomputing"},{"issue":"3","key":"2026082113304844600_bib38","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1007\/s00422-023-00963-y","article-title":"Self-organizing maps on \u201cwhat-where\u201d codes towards fully unsupervised classification","volume":"117","author":"Sa-Couto","year":"2023","journal-title":"Biological. Cybernetics"},{"issue":"3","key":"2026082113304844600_bib39","doi-asserted-by":"publisher","first-page":"411","DOI":"10.1109\/TPAMI.2007.56","article-title":"Robust object recognition with cortex-like mechanisms","volume":"29","author":"Serre","year":"2007","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"2026082113304844600_bib40","first-page":"994","article-title":"Object recognition with features inspired by visual cortex","volume-title":"Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition","author":"Serre","year":"2005"},{"issue":"3972","key":"2026082113304844600_bib41","doi-asserted-by":"publisher","first-page":"701","DOI":"10.1126\/science.171.3972.701","article-title":"Mental rotation of three-dimensional objects","volume":"171","author":"Shepard","year":"1971","journal-title":"Science"},{"key":"2026082113304844600_bib42","article-title":"Efficient pattern recognition using a new transformation distance","volume-title":"Advances in neural information processing systems","author":"Simard","year":"1992"},{"key":"2026082113304844600_bib43","doi-asserted-by":"publisher","first-page":"93","DOI":"10.3389\/fncom.2014.00093","article-title":"Hierarchical representation of shapes in visual cortex\u2014from localized features to figural shape segregation","volume":"8","author":"Tschechne","year":"2014","journal-title":"Frontiers in Computational Neuroscience"},{"issue":"3","key":"2026082113304844600_bib44","doi-asserted-by":"publisher","first-page":"193","DOI":"10.1016\/0010-0277(89)90036-X","article-title":"Aligning pictorial descriptions: An approach to object recognition","volume":"32","author":"Ullman","year":"1989","journal-title":"Cognition"},{"key":"2026082113304844600_bib45","doi-asserted-by":"publisher","first-page":"109819","DOI":"10.1016\/j.patcog.2023.109819","article-title":"Haar wavelet down-sampling: A simple but effective downsampling module for semantic segmentation","volume":"143","author":"Xu","year":"2023","journal-title":"Pattern Recognition"}],"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/38\/9\/1611\/2618232\/neco.a.1563.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/38\/9\/1611\/2618232\/neco.a.1563.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T17:30:56Z","timestamp":1787333456000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/38\/9\/1611\/137976\/Shape-Matters-Few-Shot-Object-Classification-From"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,21]]},"references-count":45,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,8,21]]},"published-print":{"date-parts":[[2026,8,21]]}},"URL":"https:\/\/doi.org\/10.1162\/neco.a.1563","relation":{},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"value":"0899-7667","type":"print"},{"value":"1530-888X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,9]]},"published":{"date-parts":[[2026,8,21]]}}}