Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations

Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations. Lee, H., Grosse, R., Ranganath, R., & Ng, A., Y. Proceedings of the 26th International Conference On Machine Learning, ICML 2009, 2009.

Paper doi abstract bibtex

There has been much interest in unsupervised learning of hierarchical generative models such as deep belief networks. Scaling such models to full-sized, high-dimensional images remains a difficult problem. To address this problem, we present the convolutional deep belief network, a hierarchical generative model which scales to realistic image sizes. This model is translation-invariant and supports efficient bottom-up and top-down probabilistic inference. Key to our approach is probabilistic max-pooling, a novel technique which shrinks the representations of higher layers in a probabilistically sound way. Our experiments show that the algorithm learns useful high-level visual features, such as object parts, from unlabeled images of objects and natural scenes. We demonstrate excellent performance on several visual recognition tasks and show that our model can perform hierarchical (bottom-up and top-down) inference over full-sized images.

@article{
 title = {Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations},
 type = {article},
 year = {2009},
 pages = {609-616},
 id = {f15f6d36-e677-362b-8eb5-664cfb0925b9},
 created = {2021-07-12T14:15:35.034Z},
 file_attached = {true},
 profile_id = {ad172e55-c0e8-3aa4-8465-09fac4d5f5c8},
 group_id = {1ff583c0-be37-34fa-9c04-73c69437d354},
 last_modified = {2021-07-12T14:16:34.993Z},
 read = {false},
 starred = {false},
 authored = {false},
 confirmed = {true},
 hidden = {false},
 folder_uuids = {85ed9c29-c272-40dc-a01a-f912101de83a},
 private_publication = {false},
 abstract = {There has been much interest in unsupervised learning of hierarchical generative models such as deep belief networks. Scaling such models to full-sized, high-dimensional images remains a difficult problem. To address this problem, we present the convolutional deep belief network, a hierarchical generative model which scales to realistic image sizes. This model is translation-invariant and supports efficient bottom-up and top-down probabilistic inference. Key to our approach is probabilistic max-pooling, a novel technique which shrinks the representations of higher layers in a probabilistically sound way. Our experiments show that the algorithm learns useful high-level visual features, such as object parts, from unlabeled images of objects and natural scenes. We demonstrate excellent performance on several visual recognition tasks and show that our model can perform hierarchical (bottom-up and top-down) inference over full-sized images.},
 bibtype = {article},
 author = {Lee, Honglak and Grosse, Roger and Ranganath, Rajesh and Ng, Andrew Y.},
 doi = {10.1145/1553374.1553453},
 journal = {Proceedings of the 26th International Conference On Machine Learning, ICML 2009},
 number = {November}
}

Downloads: 0

{"_id":"4E5YRAhdZPS54gybA","bibbaseid":"lee-grosse-ranganath-ng-convolutionaldeepbeliefnetworksforscalableunsupervisedlearningofhierarchicalrepresentations-2009","downloads":0,"creationDate":"2015-09-03T07:18:47.246Z","title":"Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations","author_short":["Lee, H.","Grosse, R.","Ranganath, R.","Ng, A., Y."],"year":2009,"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","bibdata":{"title":"Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations","type":"article","year":"2009","pages":"609-616","id":"f15f6d36-e677-362b-8eb5-664cfb0925b9","created":"2021-07-12T14:15:35.034Z","file_attached":"true","profile_id":"ad172e55-c0e8-3aa4-8465-09fac4d5f5c8","group_id":"1ff583c0-be37-34fa-9c04-73c69437d354","last_modified":"2021-07-12T14:16:34.993Z","read":false,"starred":false,"authored":false,"confirmed":"true","hidden":false,"folder_uuids":"85ed9c29-c272-40dc-a01a-f912101de83a","private_publication":false,"abstract":"There has been much interest in unsupervised learning of hierarchical generative models such as deep belief networks. Scaling such models to full-sized, high-dimensional images remains a difficult problem. To address this problem, we present the convolutional deep belief network, a hierarchical generative model which scales to realistic image sizes. This model is translation-invariant and supports efficient bottom-up and top-down probabilistic inference. Key to our approach is probabilistic max-pooling, a novel technique which shrinks the representations of higher layers in a probabilistically sound way. Our experiments show that the algorithm learns useful high-level visual features, such as object parts, from unlabeled images of objects and natural scenes. We demonstrate excellent performance on several visual recognition tasks and show that our model can perform hierarchical (bottom-up and top-down) inference over full-sized images.","bibtype":"article","author":"Lee, Honglak and Grosse, Roger and Ranganath, Rajesh and Ng, Andrew Y.","doi":"10.1145/1553374.1553453","journal":"Proceedings of the 26th International Conference On Machine Learning, ICML 2009","number":"November","bibtex":"@article{\n title = {Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations},\n type = {article},\n year = {2009},\n pages = {609-616},\n id = {f15f6d36-e677-362b-8eb5-664cfb0925b9},\n created = {2021-07-12T14:15:35.034Z},\n file_attached = {true},\n profile_id = {ad172e55-c0e8-3aa4-8465-09fac4d5f5c8},\n group_id = {1ff583c0-be37-34fa-9c04-73c69437d354},\n last_modified = {2021-07-12T14:16:34.993Z},\n read = {false},\n starred = {false},\n authored = {false},\n confirmed = {true},\n hidden = {false},\n folder_uuids = {85ed9c29-c272-40dc-a01a-f912101de83a},\n private_publication = {false},\n abstract = {There has been much interest in unsupervised learning of hierarchical generative models such as deep belief networks. Scaling such models to full-sized, high-dimensional images remains a difficult problem. To address this problem, we present the convolutional deep belief network, a hierarchical generative model which scales to realistic image sizes. This model is translation-invariant and supports efficient bottom-up and top-down probabilistic inference. Key to our approach is probabilistic max-pooling, a novel technique which shrinks the representations of higher layers in a probabilistically sound way. Our experiments show that the algorithm learns useful high-level visual features, such as object parts, from unlabeled images of objects and natural scenes. We demonstrate excellent performance on several visual recognition tasks and show that our model can perform hierarchical (bottom-up and top-down) inference over full-sized images.},\n bibtype = {article},\n author = {Lee, Honglak and Grosse, Roger and Ranganath, Rajesh and Ng, Andrew Y.},\n doi = {10.1145/1553374.1553453},\n journal = {Proceedings of the 26th International Conference On Machine Learning, ICML 2009},\n number = {November}\n}","author_short":["Lee, H.","Grosse, R.","Ranganath, R.","Ng, A., Y."],"urls":{"Paper":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c/file/693a0fe2-4c91-888b-7eb0-772effef28c2/Convolutional_deep_belief_networks_for_scalable_un.pdf.pdf"},"biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","bibbaseid":"lee-grosse-ranganath-ng-convolutionaldeepbeliefnetworksforscalableunsupervisedlearningofhierarchicalrepresentations-2009","role":"author","metadata":{"authorlinks":{}},"downloads":0},"search_terms":["convolutional","deep","belief","networks","scalable","unsupervised","learning","hierarchical","representations","lee","grosse","ranganath","ng"],"keywords":[],"authorIDs":[],"dataSources":["EmYaiv9TCHbg7caTW","ya2CyA73rpZseyrZ8","XFrKPG99s5t3W7xuW","2252seNhipfTmjEBQ"]}