{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T05:58:54Z","timestamp":1772085534960,"version":"3.50.1"},"reference-count":75,"publisher":"IOP Publishing","issue":"4","license":[{"start":{"date-parts":[[2022,10,21]],"date-time":"2022-10-21T00:00:00Z","timestamp":1666310400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2022,10,21]],"date-time":"2022-10-21T00:00:00Z","timestamp":1666310400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/100010686","name":"H2020 European Institute of Innovation and Technology","doi-asserted-by":"crossref","award":["732642"],"award-info":[{"award-number":["732642"]}],"id":[{"id":"10.13039\/100010686","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Neuromorph. Comput. Eng."],"published-print":{"date-parts":[[2022,12,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Event-based cameras encode changes in a visual scene with high temporal precision and low power consumption, generating millions of events per second in the process. Current event-based processing algorithms do not scale well in terms of runtime and computational resources when applied to a large amount of data. This problem is further exacerbated by the development of high spatial resolution vision sensors. We introduce a fast and computationally efficient clustering algorithm that is particularly designed for dealing with large event-based datasets. The approach is based on the expectation-maximization (EM) algorithm and relies on a stochastic approximation of the E-step over a truncated space to reduce the computational burden and speed up the learning process. We evaluate the quality, complexity, and stability of the clustering algorithm on a variety of large event-based datasets, and then validate our approach with a classification task. The proposed algorithm is significantly faster than standard k-means and reduces computational demands by two to three orders of magnitude while being more stable, interpretable, and close to the state of the art in terms of classification accuracy.<\/jats:p>","DOI":"10.1088\/2634-4386\/ac970d","type":"journal-article","created":{"date-parts":[[2022,10,3]],"date-time":"2022-10-03T22:17:50Z","timestamp":1664835470000},"page":"044004","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Efficient spatio-temporal feature clustering for large event-based datasets"],"prefix":"10.1088","volume":"2","author":[{"given":"Omar","family":"Oubari","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2517-5782","authenticated-orcid":true,"given":"Georgios","family":"Exarchakis","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0619-3555","authenticated-orcid":true,"given":"Gregor","family":"Lenz","sequence":"additional","affiliation":[]},{"given":"Ryad","family":"Benosman","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0030-6574","authenticated-orcid":true,"given":"Sio-Hoi","family":"Ieng","sequence":"additional","affiliation":[]}],"member":"266","published-online":{"date-parts":[[2022,10,21]]},"reference":[{"key":"nceac970dbib1","doi-asserted-by":"publisher","first-page":"566","DOI":"10.1109\/jssc.2007.914337","article-title":"A 128 \u00d7 128 120 dB 15 us latency asynchronous temporal contrast vision sensor","volume":"43","author":"Lichtsteiner","year":"2008","journal-title":"IEEE J. Solid-State Circuits"},{"key":"nceac970dbib2","doi-asserted-by":"publisher","first-page":"259","DOI":"10.1109\/jssc.2010.2085952","article-title":"A QVGA 143 dB dynamic range frame-free PWM image sensor with lossless pixel-level video compression and time-domain CDS","volume":"46","author":"Posch","year":"2011","journal-title":"IEEE J. Solid-State Circuits"},{"key":"nceac970dbib3","doi-asserted-by":"publisher","first-page":"174","DOI":"10.1007\/978-3-642-04667-4_18","article-title":"Bio-inspired stereo vision system with silicon retina imagers","author":"Kogler","year":"2009"},{"key":"nceac970dbib4","doi-asserted-by":"publisher","first-page":"3857","DOI":"10.1109\/cvpr.2019.00398","article-title":"Events-to-video: bringing modern computer vision to event cameras","author":"Rebecq","year":"2019"},{"key":"nceac970dbib5","doi-asserted-by":"publisher","first-page":"601","DOI":"10.1007\/s11263-019-01209-w","article-title":"Eklt: asynchronous photometric feature tracking using events and frames","volume":"128","author":"Gehrig","year":"2020","journal-title":"Int. J. Comput. Vis."},{"key":"nceac970dbib6","doi-asserted-by":"publisher","first-page":"407","DOI":"10.1109\/tnnls.2013.2273537","article-title":"Event-based visual flow","volume":"25","author":"Benosman","year":"2014","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"nceac970dbib7","doi-asserted-by":"publisher","first-page":"1346","DOI":"10.1109\/TPAMI.2016.2574707","article-title":"Hots: a hierarchy of event-based time-surfaces for pattern recognition","volume":"39","author":"Lagorce","year":"2016","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"nceac970dbib8","article-title":"Event-based vision: a survey","author":"Gallego","year":"2019"},{"key":"nceac970dbib9","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"LeCun","year":"1998","journal-title":"Proc. IEEE"},{"key":"nceac970dbib10","first-page":"1097","article-title":"Imagenet classification with deep convolutional neural networks","author":"Krizhevsky","year":"2012"},{"key":"nceac970dbib11","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ebccsp.2016.7605086","article-title":"Feature detection and tracking with the dynamic and active-pixel vision sensor (davis)","author":"Tedaldi","year":"2016"},{"key":"nceac970dbib12","doi-asserted-by":"publisher","first-page":"5419","DOI":"10.1109\/cvpr.2018.00568","article-title":"Event-based vision meets deep learning on steering prediction for self-driving cars","author":"Maqueda","year":"2018"},{"key":"nceac970dbib13","doi-asserted-by":"publisher","first-page":"1381","DOI":"10.1007\/s11263-018-1106-2","article-title":"Real-time intensity-image reconstruction for event cameras using manifold regularisation","volume":"126","author":"Munda","year":"2018","journal-title":"Int. J. Comput. Vis."},{"key":"nceac970dbib14","doi-asserted-by":"publisher","first-page":"156","DOI":"10.1109\/wacv45572.2020.9093366","article-title":"Fast image reconstruction with an event camera","author":"Scheerlinck","year":"2020"},{"key":"nceac970dbib15","doi-asserted-by":"publisher","first-page":"1826","DOI":"10.1109\/wacv.2019.00199","article-title":"Space-time event clouds for gesture recognition: from rgb cameras to event cameras","author":"Wang","year":"2019"},{"key":"nceac970dbib16","doi-asserted-by":"publisher","first-page":"3887","DOI":"10.1109\/cvpr.2019.00401","article-title":"Eventnet: asynchronous recursive event processing","author":"Sekikawa","year":"2019"},{"key":"nceac970dbib17","first-page":"652","article-title":"Pointnet: deep learning on point sets for 3D classification and segmentation","author":"Qi","year":"2017"},{"key":"nceac970dbib18","doi-asserted-by":"publisher","first-page":"1659","DOI":"10.1016\/s0893-6080(97)00011-7","article-title":"Networks of spiking neurons: the third generation of neural network models","volume":"10","author":"Maass","year":"1997","journal-title":"Neural Netw."},{"key":"nceac970dbib19","doi-asserted-by":"publisher","first-page":"508","DOI":"10.3389\/fnins.2016.00508","article-title":"Training deep spiking neural networks using backpropagation","volume":"10","author":"Lee","year":"2016","journal-title":"Front. Neurosci."},{"key":"nceac970dbib20","first-page":"1412","article-title":"Slayer: spike layer error reassignment in time","author":"Shrestha","year":"2018"},{"key":"nceac970dbib21","doi-asserted-by":"publisher","first-page":"424","DOI":"10.3389\/fnins.2020.00424","article-title":"Synaptic plasticity dynamics for deep continuous local learning (decolle)","volume":"14","author":"Kaiser","year":"2020","journal-title":"Front. Neurosci."},{"key":"nceac970dbib22","doi-asserted-by":"publisher","first-page":"2050027","DOI":"10.1142\/s0129065720500276","article-title":"S4nn: temporal backpropagation for spiking neural networks with one spike per neuron","volume":"30","author":"Kheradpisheh","year":"2020","journal-title":"Int. J. Neural Syst."},{"key":"nceac970dbib23","doi-asserted-by":"publisher","first-page":"54","DOI":"10.1007\/s11263-014-0788-3","article-title":"Spiking deep convolutional neural networks for energy-efficient object recognition","volume":"113","author":"Cao","year":"2015","journal-title":"Int. J. Comput. Vis."},{"key":"nceac970dbib24","doi-asserted-by":"publisher","first-page":"682","DOI":"10.3389\/fnins.2017.00682","article-title":"Conversion of continuous-valued deep networks to efficient event-driven networks for image classification","volume":"11","author":"Rueckauer","year":"2017","journal-title":"Front. Neurosci."},{"key":"nceac970dbib25","doi-asserted-by":"publisher","first-page":"1319","DOI":"10.1609\/aaai.v33i01.33011319","article-title":"Tdsnn: from deep neural networks to deep spike neural networks with temporal-coding","volume":"vol 33","author":"Zhang","year":"2019"},{"key":"nceac970dbib26","doi-asserted-by":"publisher","first-page":"119","DOI":"10.3389\/fnins.2020.00119","article-title":"Enabling spike-based backpropagation for training deep neural network architectures","volume":"14","author":"Lee","year":"2020","journal-title":"Front. Neurosci."},{"key":"nceac970dbib27","doi-asserted-by":"publisher","first-page":"439","DOI":"10.3389\/fnins.2020.00439","article-title":"Efficient processing of spatio-temporal data streams with spiking neural networks","volume":"14","author":"Kugele","year":"2020","journal-title":"Front. Neurosci."},{"key":"nceac970dbib28","doi-asserted-by":"publisher","first-page":"763","DOI":"10.1007\/978-3-540-78293-3_18","article-title":"Neural systems engineering","author":"Furber","year":"2008"},{"key":"nceac970dbib29","doi-asserted-by":"publisher","first-page":"82","DOI":"10.1109\/mm.2018.112130359","article-title":"Loihi: a neuromorphic manycore processor with on-chip learning","volume":"38","author":"Davies","year":"2018","journal-title":"IEEE Micro"},{"key":"nceac970dbib30","doi-asserted-by":"publisher","DOI":"10.1117\/12.2305933","article-title":"A sparse coding multi-scale precise-timing machine learning algorithm for neuromorphic event-based sensors","volume":"vol 10639","author":"Haessig","year":"2018"},{"key":"nceac970dbib31","article-title":"Event-based feature extraction using adaptive selection thresholds","author":"Afshar","year":"2019"},{"key":"nceac970dbib32","doi-asserted-by":"publisher","first-page":"1731","DOI":"10.1109\/cvpr.2018.00186","article-title":"Hats: histograms of averaged time surfaces for robust event-based object classification","author":"Sironi","year":"2018"},{"key":"nceac970dbib33","article-title":"Learning to detect objects with a 1 megapixel event camera","author":"Perot","year":"2020"},{"key":"nceac970dbib34","doi-asserted-by":"publisher","DOI":"10.1109\/cvprw.2019.00214","article-title":"Live demonstration: Celex-v: a 1\u00a0m pixel multi-mode event-based sensor","author":"Chen","year":"2019"},{"key":"nceac970dbib35","doi-asserted-by":"publisher","first-page":"47","DOI":"10.1109\/cidu.2012.6382200","article-title":"Introduction to astroml: machine learning for astrophysics","author":"VanderPlas","year":"2012"},{"key":"nceac970dbib36","first-page":"119","article-title":"The hand mouse: Gmm hand-color classification and mean shift tracking","author":"Kurata","year":"2001"},{"key":"nceac970dbib37","first-page":"2224","article-title":"Modified gmm background modeling and optical flow for detection of moving objects","volume":"vol 3","author":"Zhou","year":"2005"},{"key":"nceac970dbib38","first-page":"I","article-title":"Robust speech recognition in additive and channel noise environments using gmm and em algorithm","volume":"vol 1","author":"Fujimoto","year":"2004"},{"key":"nceac970dbib39","first-page":"246","article-title":"Adaptive background mixture models for real-time tracking","volume":"vol 2","author":"Stauffer","year":"1999"},{"key":"nceac970dbib40","article-title":"Can clustering scale sublinearly with its clusters? A variational em acceleration of gmms and k-means","author":"Forster","year":"2017"},{"key":"nceac970dbib41","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3133763","article-title":"A variational em acceleration for efficient clustering at very large scales","author":"Hirschberger","year":"2021","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"nceac970dbib42","doi-asserted-by":"publisher","first-page":"2402","DOI":"10.1109\/TPAMI.2017.2769655","article-title":"Event-based, 6-dof\u00a0camera tracking from photometric depth maps","volume":"40","author":"Gallego","year":"2017","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"nceac970dbib43","article-title":"k-means++: the advantages of careful seeding","author":"Arthur","year":"2006"},{"key":"nceac970dbib44","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v30i1.10259","article-title":"Approximate k-means++ in sublinear time","author":"Bachem","year":"2016"},{"key":"nceac970dbib45","first-page":"55","article-title":"Fast and provably good seedings for k-means","author":"Bachem","year":"2016"},{"key":"nceac970dbib46","first-page":"291","article-title":"On coresets for k-means and k-median clustering","author":"Har-Peled","year":"2004"},{"key":"nceac970dbib47","doi-asserted-by":"publisher","first-page":"1119","DOI":"10.1145\/3219819.3219973","article-title":"Scalable k-means clustering via lightweight coresets","author":"Bachem","year":"2018"},{"key":"nceac970dbib48","first-page":"2855","article-title":"Expectation truncation and the benefits of preselection in training generative models","volume":"11","author":"L\u00fccke","year":"2010","journal-title":"J. Mach. Learn. Res."},{"key":"nceac970dbib49","first-page":"243","article-title":"What are the invariant occlusive components of image patches? A probabilistic generative approach","author":"Dai","year":"2013"},{"key":"nceac970dbib50","doi-asserted-by":"publisher","first-page":"2979","DOI":"10.1162\/neco_a_01015","article-title":"Discrete sparse coding","volume":"29","author":"Exarchakis","year":"2017","journal-title":"Neural Comput."},{"key":"nceac970dbib51","doi-asserted-by":"publisher","first-page":"3769","DOI":"10.1109\/ijcnn.2017.7966331","article-title":"Truncated variational em for semi-supervised neural simpletrons","author":"Forster","year":"2017"},{"key":"nceac970dbib52","doi-asserted-by":"publisher","first-page":"467","DOI":"10.1007\/978-3-319-93764-9_43","article-title":"Truncated variational sampling for \u2018black box\u2019 optimization of generative models","author":"L\u00fccke","year":"2018"},{"key":"nceac970dbib53","doi-asserted-by":"publisher","first-page":"204","DOI":"10.1007\/978-3-642-28551-6_26","article-title":"Ternary sparse coding","author":"Exarchakis","year":"2012"},{"key":"nceac970dbib54","first-page":"4116","article-title":"Local maxima in the likelihood of Gaussian mixture models: structural results and algorithmic consequences","author":"Jin","year":"2016"},{"key":"nceac970dbib55","doi-asserted-by":"publisher","first-page":"481","DOI":"10.3389\/fnins.2015.00481","article-title":"Poker-dvs and mnist-dvs their history, how they were made, and other details","volume":"9","author":"Serrano-Gotarredona","year":"2015","journal-title":"Front. Neurosci."},{"key":"nceac970dbib56","doi-asserted-by":"publisher","first-page":"437","DOI":"10.3389\/fnins.2015.00437","article-title":"Converting static image datasets to spiking neuromorphic datasets using saccades","volume":"9","author":"Orchard","year":"2015","journal-title":"Front. Neurosci."},{"key":"nceac970dbib57","doi-asserted-by":"publisher","first-page":"7243","DOI":"10.1109\/cvpr.2017.781","article-title":"A low power, fully event-based gesture recognition system","author":"Amir","year":"2017"},{"key":"nceac970dbib58","doi-asserted-by":"publisher","first-page":"705","DOI":"10.1007\/978-3-030-01267-0_43","article-title":"Hgmr: hierarchical Gaussian mixtures for adaptive 3D registration","author":"Eckart","year":"2018"},{"key":"nceac970dbib59","doi-asserted-by":"publisher","first-page":"1397","DOI":"10.1109\/TPAMI.2017.2717829","article-title":"Joint alignment of multiple point sets with batch and incremental expectation-maximization","volume":"40","author":"Evangelidis","year":"2017","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"nceac970dbib60","article-title":"Tonic: event-based datasets and transformations","author":"Lenz","year":"2021"},{"key":"nceac970dbib61","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1111\/j.2517-6161.1977.tb01600.x","article-title":"Maximum likelihood from incomplete data via the em algorithm","volume":"39","author":"Dempster","year":"1977","journal-title":"J. R. Stat. Soc. B"},{"key":"nceac970dbib62","author":"Bishop","year":"2006"},{"key":"nceac970dbib63","doi-asserted-by":"publisher","first-page":"355","DOI":"10.1007\/978-94-011-5014-9_12","article-title":"A view of the em algorithm that justifies incremental, sparse, and other variants","author":"Neal","year":"1998"},{"key":"nceac970dbib64","first-page":"2142","article-title":"Scalable training of mixture models via coresets","author":"Feldman","year":"2011"},{"key":"nceac970dbib65","first-page":"185","article-title":"Eventbuilder: real-time multimedia event summarization by visualizing social media","author":"Shah","year":"2015"},{"key":"nceac970dbib66","doi-asserted-by":"publisher","first-page":"102","DOI":"10.1016\/j.knosys.2016.05.022","article-title":"Leveraging multimodal information for event summarization and concept-level sentiment analysis","volume":"108","author":"Shah","year":"2016","journal-title":"Knowl.-Based Syst."},{"key":"nceac970dbib67","first-page":"5885","article-title":"Training Gaussian mixture models at scale via coresets","volume":"18","author":"Lucic","year":"2017","journal-title":"J. Mach. Learn. Res."},{"key":"nceac970dbib68","doi-asserted-by":"publisher","first-page":"761","DOI":"10.1016\/j.patcog.2005.09.012","article-title":"Iterative shrinking method for clustering problems","volume":"39","author":"Fr\u00e4nti","year":"2006","journal-title":"Pattern Recognit."},{"key":"nceac970dbib69","doi-asserted-by":"publisher","first-page":"469","DOI":"10.1162\/089976603762553004","article-title":"Efficient greedy learning of Gaussian mixture models","volume":"15","author":"Verbeek","year":"2003","journal-title":"Neural Comput."},{"key":"nceac970dbib70","doi-asserted-by":"publisher","first-page":"235","DOI":"10.1561\/2200000008","article-title":"Clustering stability: an overview","volume":"2","author":"von Luxor","year":"2010","journal-title":"Found. Trends"},{"key":"nceac970dbib71","doi-asserted-by":"publisher","first-page":"377","DOI":"10.1007\/978-3-540-30541-5_47","article-title":"Spatial histogram features for face detection in color images","author":"Zhang","year":"2004"},{"key":"nceac970dbib72","doi-asserted-by":"publisher","first-page":"45974","DOI":"10.1109\/access.2020.2978109","article-title":"An end-to-end broad learning system for event-based object classification","volume":"8","author":"Gao","year":"2020","journal-title":"IEEE Access"},{"key":"nceac970dbib73","doi-asserted-by":"publisher","first-page":"1177","DOI":"10.1145\/1772690.1772862","article-title":"Web-scale k-means clustering","author":"Sculley","year":"2010"},{"key":"nceac970dbib74","doi-asserted-by":"publisher","first-page":"2028","DOI":"10.1109\/tpami.2015.2392947","article-title":"Hfirst: a temporal approach to object recognition","volume":"37","author":"Orchard","year":"2015","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"nceac970dbib75","doi-asserted-by":"publisher","first-page":"770","DOI":"10.1109\/cvpr.2016.90","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"}],"container-title":["Neuromorphic Computing and Engineering"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,10,20]],"date-time":"2022-10-20T07:59:42Z","timestamp":1666252782000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2634-4386\/ac970d"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,10,21]]},"references-count":75,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2022,10,21]]},"published-print":{"date-parts":[[2022,12,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2634-4386\/ac970d","relation":{},"ISSN":["2634-4386"],"issn-type":[{"value":"2634-4386","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,21]]},"assertion":[{"value":"Efficient spatio-temporal feature clustering for large event-based datasets","name":"article_title","label":"Article Title"},{"value":"Neuromorphic Computing and Engineering","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2022 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2022-01-17","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2022-10-03","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2022-10-21","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}