Open AccessDOI
Person Identification in Webcam Images: An Application of Semi-Supervised Learning
Maria-Florina Balcan,Avrim Blum,Patrick Pakyan Choi,John Lafferty,Brian Pantano,Mugizi Robert Rwebangira,Xiaojin Zhu +6 more
Reads0
Chats0
TLDR
The person identification task is posed as a graph-based semi-supervised learning problem, where only a few training images are labeled and the importance of domain knowledge in graph construction is discussed, and experiments are presented that clearly show the advantage of semi- supervised learning over standard supervised learning.Abstract:
An application of semi-supervised learning is
made to the problem of person identification in
low quality webcam images. Using a set of images
of ten people collected over a period of four
months, the person identification task is posed
as a graph-based semi-supervised learning problem,
where only a few training images are labeled.
The importance of domain knowledge
in graph construction is discussed, and experiments
are presented that clearly show the advantage
of semi-supervised learning over standard
supervised learning. The data used in the study
is available to the research community to encourage
further investigation of this problem.read more
Citations
More filters
Book
Introduction to Semi-Supervised Learning
TL;DR: This introductory book presents some popular semi-supervised learning models, including self-training, mixture models, co-training and multiview learning, graph-based methods, and semi- supervised support vector machines, and discusses their basic mathematical formulation.
Journal ArticleDOI
Label Propagation through Linear Neighborhoods
Fei Wang,Changshui Zhang +1 more
TL;DR: A novel graph-based semi supervised learning approach is proposed based on a linear neighborhood model, which assumes that each data point can be linearly reconstructed from its neighborhood, and can propagate the labels from the labeled points to the whole data set using these linear neighborhoods with sufficient smoothness.
Journal ArticleDOI
Unified Video Annotation via Multigraph Learning
TL;DR: This paper shows that various crucial factors in video annotation, including multiple modalities, multiple distance functions, and temporal consistency, all correspond to different relationships among video units, and hence they can be represented by different graphs, and proposes optimized multigraph-based semi-supervised learning (OMG-SSL), which aims to simultaneously tackle these difficulties in a unified scheme.
Proceedings ArticleDOI
Label propagation through linear neighborhoods
Fei Wang,Changshui Zhang +1 more
TL;DR: A novel graph-based semi supervised learning approach is proposed based on a linear neighborhood model, which assumes that each data point can be linearly reconstructed from its neighborhood, and can propagate the labels from the labeled points to the whole data set using these linear neighborhoods with sufficient smoothness.
References
More filters
Book
Fundamentals of digital image processing
TL;DR: This chapter discusses two Dimensional Systems and Mathematical Preliminaries and their applications in Image Analysis and Computer Vision, as well as image reconstruction from Projections and image enhancement.
Journal ArticleDOI
Face recognition: A literature survey
TL;DR: In this paper, the authors provide an up-to-date critical survey of still-and video-based face recognition research, and provide some insights into the studies of machine recognition of faces.
Proceedings ArticleDOI
Combining labeled and unlabeled data with co-training
Avrim Blum,Tom M. Mitchell +1 more
TL;DR: A PAC-style analysis is provided for a problem setting motivated by the task of learning to classify web pages, in which the description of each example can be partitioned into two distinct views, to allow inexpensive unlabeled data to augment, a much smaller set of labeled examples.
Proceedings Article
Semi-supervised learning using Gaussian fields and harmonic functions
TL;DR: An approach to semi-supervised learning is proposed that is based on a Gaussian random field model, and methods to incorporate class priors and the predictions of classifiers obtained by supervised learning are discussed.
Support Vector Machines for Classification and Regression
TL;DR: The Structural Risk Minimization (SRM) as discussed by the authors principle has been shown to be superior to traditional empirical risk minimization (ERM) principle employed by conventional neural networks, as opposed to ERM which minimizes the error on the training data.