B. J. Culpepper, Jascha Narain Sohl-Dickstein, B. Olshausen
{"title":"用因子分解法建立更好的图像概率模型","authors":"B. J. Culpepper, Jascha Narain Sohl-Dickstein, B. Olshausen","doi":"10.1109/ICCV.2011.6126473","DOIUrl":null,"url":null,"abstract":"We describe a directed bilinear model that learns higher-order groupings among features of natural images. The model represents images in terms of two sets of latent variables: one set of variables represents which feature groups are active, while the other specifies the relative activity within groups. Such a factorized representation is beneficial because it is stable in response to small variations in the placement of features while still preserving information about relative spatial relationships. When trained on MNIST digits, the resulting representation provides state of the art performance in classification using a simple classifier. When trained on natural images, the model learns to group features according to proximity in position, orientation, and scale. The model achieves high log-likelihood (−94 nats), surpassing the current state of the art for natural images achievable with an mcRBM model.","PeriodicalId":6391,"journal":{"name":"2011 International Conference on Computer Vision","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2011-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"14","resultStr":"{\"title\":\"Building a better probabilistic model of images by factorization\",\"authors\":\"B. J. Culpepper, Jascha Narain Sohl-Dickstein, B. Olshausen\",\"doi\":\"10.1109/ICCV.2011.6126473\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We describe a directed bilinear model that learns higher-order groupings among features of natural images. The model represents images in terms of two sets of latent variables: one set of variables represents which feature groups are active, while the other specifies the relative activity within groups. Such a factorized representation is beneficial because it is stable in response to small variations in the placement of features while still preserving information about relative spatial relationships. When trained on MNIST digits, the resulting representation provides state of the art performance in classification using a simple classifier. When trained on natural images, the model learns to group features according to proximity in position, orientation, and scale. The model achieves high log-likelihood (−94 nats), surpassing the current state of the art for natural images achievable with an mcRBM model.\",\"PeriodicalId\":6391,\"journal\":{\"name\":\"2011 International Conference on Computer Vision\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2011-11-06\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"14\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2011 International Conference on Computer Vision\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICCV.2011.6126473\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2011 International Conference on Computer Vision","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCV.2011.6126473","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Building a better probabilistic model of images by factorization
We describe a directed bilinear model that learns higher-order groupings among features of natural images. The model represents images in terms of two sets of latent variables: one set of variables represents which feature groups are active, while the other specifies the relative activity within groups. Such a factorized representation is beneficial because it is stable in response to small variations in the placement of features while still preserving information about relative spatial relationships. When trained on MNIST digits, the resulting representation provides state of the art performance in classification using a simple classifier. When trained on natural images, the model learns to group features according to proximity in position, orientation, and scale. The model achieves high log-likelihood (−94 nats), surpassing the current state of the art for natural images achievable with an mcRBM model.