Unsupervised learning of overcomplete face descriptors.

Its applications to face verification and retrieval outperform existing state-of-the-art methods in accuracy while retaining high efficiency. ... This paper addresses both aspects by an ensemble metric learning approach that consists of sparse block diagonal metric ensembling and join- t metric learning as two consecutive steps. ... unsupervised metric learning method. ...

arXiv:1212.6094v1 fatcat:2tzhrsdzfjdgbn3tpt2grjp3f4

Experiments show that our supervised dictionary improves the performance of the proposed model significantly over the unsupervised dictionary, leading to state-of-the-art performance on diverse image databases ... In this paper, we propose a novel supervised hierarchical sparse coding model based on local image descriptors for classification tasks. ... Acknowledgement The main part of this work was done when the first author was a summer intern at NEC Laboratories America in Cupertino, CA. The work is also supported in part by the U.S. ...

doi:10.1109/cvpr.2010.5539958 dblp:conf/cvpr/YangYH10 fatcat:36d2xms3ibcdtiituf6zdonmq4

A 100K-dim feature, based on a single-type Local Binary Pattern (LBP) descriptor, can achieve significant improvements over both its low-dimensional version and the state-of-the-art. ... This prevents further exploration of the use of a highdimensional feature. In this paper, we study the performance of a highdimensional feature. ... High-dimensional feature with unsupervised learning In this experiment, we study the impact of high dimensionality under the unsupervised setting. ...

doi:10.1109/cvpr.2013.389 dblp:conf/cvpr/ChenCWS13 fatcat:4u4lnjypo5h5xfoxgcprlbqyjq

Training dictionaries such that they represent each class of signals with minimal loss is called Dictionary Learning (DL). ... This chapter presents a review of statistical techniques and their application to learning discriminative dictionaries. ... In [43] , group sparse coding with HoG feature descriptors is used to achieve good results on face recognition. ...

arXiv:2111.09027v1 fatcat:zzdz2m3rvfauxjw4eu2mgpk5mu

Open Access

This paper explores the effectiveness of sparse representations obtained by learning a set of overcomplete basis (dictionary) in the context of action recognition in videos. ... An overcomplete dictionary is constructed using a set of spatio-temporal descriptors (extracted from the video sequences) in such a way that each descriptor is represented by some linear combination of ... THE PROPOSED APPROACH Our approach broadly consists of four stages : computation of the spatio-temporal motion descriptors, di-mensionality reduction of the descriptors, learning overcomplete dictionaries ...

doi:10.1109/tpami.2011.253 pmid:22745001 fatcat:6ime7uvamfgq3jg5dz4nrj7cya

Moreover, we develop a new distance metric learning method for face verification called Pairwise-constrained Multiple Metric Learning (PMML) to effectively integrate the face region descriptors of all ... Spatial-Temporal Face Region Descriptor, STFRD) for images (resp. videos). ... Face Region Descriptor and PMML Our approach consists of two key components: face region descriptors and the learning algorithm PMML. ...

doi:10.1109/cvpr.2013.456 dblp:conf/cvpr/CuiLXSC13 fatcat:iyikihhqmvfz3jgfejsbmg26km

From a collection of realworld images, we learn an overcomplete dictionary to represent visual prior. ... The prior knowledge of objects is generic, and the training image set does not necessarily contain any observation of the target object. ... Wang was a visiting student at the University of California at Merced. ...

doi:10.1109/tip.2012.2190085 pmid:22491081 fatcat:obwthxkxqvag5eo6rwtqsaqpp4

The locally encoded sparse vectors are pooled to form regional descriptors, where each descriptor covers a relatively large portion of the face. ... Most of the related literature focuses on holistic descriptors in closed-set identification applications. ... Each sparse encoding approach learns a dictionary (or model), D ∈ R d×N , using an unsupervised learning algorithm, where each column d i ∈ R d of the dictionary is an atom. ...

doi:10.1109/ijcnn.2012.6252611 dblp:conf/ijcnn/WongHSL12 fatcat:6xuf7p62zfgkpj3ophe3xuoe5e

During this work, multiscale DoG is first performed to come up with overcomplete face illustration, and then three native descriptors specifically HOG, GLOH and SIFT area unit applied to construct the ... In this work, we have a tendency to explore the basis of the principle of some popular existing descriptors for VIS-NIR matching and more introduce our projected descriptor besides illumination invariant ...

doi:10.33980/ajabs.2019.v07i01.002 fatcat:uoqwawpoanh43eo4stdlglnyfy

Open Access

22 0274-6638/16©2016IEEE ieee Geoscience and remote sensinG maGazine jUNE 2016 d eep-learning (DL) algorithms, which learn the representative and discriminative features in a hierarchical manner from the ... data, have recently become a hotspot in the machine-learning area and have been introduced into the geoscience and remote sensing (RS) community for RS big data analysis. ... sParse CodinG Sparse coding is a type of unsupervised method for learning sets of overcomplete bases to represent data efficiently to find a set of basis vectors i z such that we can represent an input ...

doi:10.1109/mgrs.2016.2540798 fatcat:yuqd4e2bm5bijbrssdxmu2zxcq

Conclusions: (1) Instead of constructing Gabor functions or Gaussian pyramids filters for feature extraction as traditional attention models do, we learn overcomplete basis sets from natural images to ... Experiments show that given the proper layer size and a robust combination strategy, the learned overcomplete basis set outperforms a complete set and Gabor pyramids in visual saliency detection. ... Acknowledgements The research is supported by National Nature Science Foundation of China (60902058,60975078,61105119,61273364), Beijing Natural Science Foundation (4112047) and fundamental Research Funds ...

doi:10.1186/1471-2202-13-145 pmid:23190754 pmcid:PMC3599588 fatcat:bjwe4evwavhl3n3swavspw5j2m

DOAJ

Coupling label evolution with the spatially regularized assignment flow induces a sparsifying effect that enables to learn compact label dictionaries in an unsupervised manner. ... The latter component of the approach encompasses extensions of state-of-the-art clustering approaches to manifold-valued data. ... Unsupervised learning of rotationally invariant labels from covariance descriptors. ...

arXiv:1904.10863v2 fatcat:7w4thzqvynd6xevd2w4xu22myi

Multiple Versions

In this method, dynamic texture is used as descriptors in a spatiotemporal manner to describe spatio-temporal volumes of events in videos. ... A novel unsupervised method is proposed to detect abnormal events by combining dynamic texture and sparse coding. ... Acknowledgements This work is supported by the National Natural Science Foundation of China under Grant no.61103123. ...

doi:10.14257/ijsip.2014.7.4.11 fatcat:orqcyv52ubhcfdfipu2ykol6gm

Inspired by its multi-level feature learning ability, a stacked denoising autoencoder (SDAE) is used to learn multi-level feature descriptors from a set of auxiliary images. ... Experiments on an open tracking benchmark show promising results of the proposed tracker as compared with several state-of-the-art approaches. ... GENERIC IMAGE FEATURE DESCRIPTOR In this section, we describe the unsupervised generic feature learning scheme, which is based on a stacked denoising autoencoder (SDAE). ...

doi:10.1109/icip.2014.7025169 dblp:conf/icip/ZhouXZZ14 fatcat:bcjn62tvhrgj5bv5jizbkp77v4

Our learning method mimics the human behavior of selective learning in the sense of fast memory renewal. ... In this work we present a novel approach for learning nonhomogenous textures without facing the unlearning problem. ... Any standard unsupervised learning algorithm such as expectation maximization can be used for this purpose. ...

doi:10.1109/isbi.2008.4541221 dblp:conf/isbi/LeeLS08 fatcat:f3cgrouolvht3hdexvk4d5skoy

Large Scale Strongly Supervised Ensemble Metric Learning, with Applications to Face Verification and Retrieval [article]

Preserved Fulltext

Supervised translation-invariant sparse coding

Preserved Fulltext

Blessing of Dimensionality: High-Dimensional Feature and Its Efficient Compression for Face Verification

Preserved Fulltext

Discriminative Dictionary Learning based on Statistical Methods [article]

Preserved Fulltext

Learning Sparse Representations for Human Action Recognition

Preserved Fulltext

Fusing Robust Face Region Descriptors via Multiple Metric Learning for Face Recognition in the Wild

Preserved Fulltext

Transferring Visual Prior for Online Object Tracking

Preserved Fulltext

On robust biometric identity verification via sparse encoding of faces: Holistic vs local approaches

Preserved Fulltext

Problem Definition on Face Recognition: A Review

Preserved Fulltext

Deep Learning for Remote Sensing Data: A Technical Tutorial on the State of the Art

Preserved Fulltext

A neural computational model for bottom-up attention with invariant and overcomplete representation

Preserved Fulltext

Unsupervised Assignment Flow: Label Learning on Feature Manifolds by Spatially Regularized Geometric Assignment [article]

Preserved Fulltext

Other Versions

Abnormal Event Detection in Nature Settings

Preserved Fulltext

An ensemble of deep neural networks for object tracking

Preserved Fulltext

Learning non-homogenous textures and the unlearning problem with application to drusen detection in retinal images

Preserved Fulltext