Please use this identifier to cite or link to this item:
https://hdl.handle.net/20.500.14279/8578
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Chatzis, Sotirios P. | - |
dc.contributor.author | Kosmopoulos, Dimitrios I. | - |
dc.date.accessioned | 2016-07-01T09:14:26Z | - |
dc.date.available | 2016-07-01T09:14:26Z | - |
dc.date.issued | 2015-12 | - |
dc.identifier.citation | 2015 IEEE International Conference on Computer Vision (ICCV), Santiago, 2015, pp. 2803-2811 | en_US |
dc.identifier.uri | https://hdl.handle.net/20.500.14279/8578 | - |
dc.description.abstract | Unsupervised feature learning algorithms based on convolutional formulations of independent components analysis (ICA) have been demonstrated to yield state-of-the-art results in several action recognition benchmarks. However, existing approaches do not allow for the number of latent components (features) to be automatically inferred from the data in an unsupervised manner. This is a significant disadvantage of the state-of-the-art, as it results in considerable burden imposed on researchers and practitioners, who must resort to tedious cross-validation procedures to obtain the optimal number of latent features. To resolve these issues, in this paper we introduce a convolutional nonparametric Bayesian sparse ICA architecture for overcomplete feature learning from high-dimensional data. Our method utilizes an Indian buffet process prior to facilitate inference of the appropriate number of latent features under a hybrid variational inference algorithm, scalable to massive datasets. As we show, our model can be naturally used to obtain deep unsupervised hierarchical feature extractors, by greedily stacking successive model layers, similar to existing approaches. In addition, inference for this model is completely heuristics-free, thus, it obviates the need of tedious parameter tuning, which is a major challenge most deep learning approaches are faced with. We evaluate our method on several action recognition benchmarks, and exhibit its advantages over the state-of-the-art. | en_US |
dc.format | en_US | |
dc.language.iso | en | en_US |
dc.rights | Copyright IEEE | en_US |
dc.subject | Bayes methods | en_US |
dc.subject | Convolution | en_US |
dc.subject | Feature extraction | en_US |
dc.subject | Image recognition | en_US |
dc.subject | Independent component analysis | en_US |
dc.title | A Nonparametric Bayesian Approach Toward Stacked Convolutional Independent Component Analysis | en_US |
dc.type | Conference Papers | en_US |
dc.link | http://pamitc.org/iccv15/ | en_US |
dc.collaboration | Cyprus University of Technology | en_US |
dc.collaboration | University of Patras | en_US |
dc.subject.category | Electrical Engineering - Electronic Engineering - Information Engineering | en_US |
dc.country | Cyprus | en_US |
dc.country | Greece | en_US |
dc.subject.field | Engineering and Technology | en_US |
dc.relation.conference | IEEE International Conference on Computer Vision (ICCV) | en_US |
dc.identifier.doi | 10.1109/ICCV.2015.321 | en_US |
dc.dept.handle | 123456789/134 | en |
cut.common.academicyear | 2015-2016 | en_US |
item.languageiso639-1 | en | - |
item.cerifentitytype | Publications | - |
item.openairecristype | http://purl.org/coar/resource_type/c_c94f | - |
item.fulltext | No Fulltext | - |
item.grantfulltext | none | - |
item.openairetype | conferenceObject | - |
crisitem.author.dept | Department of Electrical Engineering, Computer Engineering and Informatics | - |
crisitem.author.faculty | Faculty of Engineering and Technology | - |
crisitem.author.orcid | 0000-0002-4956-4013 | - |
crisitem.author.parentorg | Faculty of Engineering and Technology | - |
Appears in Collections: | Δημοσιεύσεις σε συνέδρια /Conference papers or poster or presentation |
CORE Recommender
SCOPUSTM
Citations
50
4
checked on Nov 9, 2023
Page view(s) 50
479
Last Week
6
6
Last month
37
37
checked on Mar 7, 2025
Google ScholarTM
Check
Altmetric
Items in KTISIS are protected by copyright, with all rights reserved, unless otherwise indicated.