68 research outputs found
Human action attribute learning from video data using low-rank representations
No full textRepresentation of human actions as a sequence of human body movements or action attributes enables the development of models for human activity recognition and summarization. We present an extension of the low-rank representation (LRR) model, termed the clustering-aware structure-constrained low-rank representation (CS-LRR) model, for unsupervised learning of human action attributes from video data. Our model is based on the union-of-subspaces (UoS) framework, and integrates spectral clustering into the LRR optimization problem for better subspace clustering results. We lay out an efficient linear alternating direction method to solve the CS-LRR optimization problem. We also introduce a hierarchical subspace clustering approach, termed hierarchical CS-LRR, to learn the attributes without the need for a priori specification of their number. By visualizing and labeling these action attributes, the hierarchical model can be used to semantically summarize long video sequences of human actions at multiple resolutions. A human action or activity can also be uniquely represented as a sequence of transitions from one action attribute to another, which can then be used for human action recognition. We demonstrate the effectiveness of the proposed model for semantic summarization and action recognition through comprehensive experiments on five real-world human action datasets.Technical Report #2020-07-00
2010 Seventh International Conference on Information Technology: New Generations
No full textIsaac Macwan (with Hassan Bajwa, Vignesh Veerapandian, and Xinghao Chen) is a contributing author, VHDL Implementation of High-Performance and Dynamically Configures Multi-port Cache Memory, pp. 1212-1216.https://digitalcommons.fairfield.edu/engineering-books/1057/thumbnail.jp
Higher order photoprotection mutants reveal the importance of ΔpH-dependent photosynthesis-control in preventing light induced damage to both photosystem II and photosystem I
No full textAlthough light is essential for photosynthesis, when in excess, it may damage the photosynthetic
apparatus, leading to a phenomenon known as photoinhibition. Photoinhibition was thought as a
light-induced damage to photosystem II; however, it is now clear that even photosystem I may become
very vulnerable to light. One main characteristic of light induced damage to photosystem II (PSII) is
the increased turnover of the reaction center protein, D1: when rate of degradation exceeds the rate of
synthesis, loss of PSII activity is observed. With respect to photosystem I (PSI), an excess of electrons,
instead of an excess of light, may be very dangerous. Plants possess a number of mechanisms able
to prevent, or limit, such damages by safe thermal dissipation of light energy (non-photochemical
quenching, NPQ), slowing-down of electron transfer through the intersystem transport chain
(photosynthesis-control, PSC) in co-operation with the Proton Gradient Regulation (PGR) proteins,
PGR5 and PGRL1, collectively called as short-term photoprotection mechanisms, and the redistribution
of light between photosystems, called state transitions (responsible of fluorescence quenching at PSII,
qT), is superimposed to these short term photoprotective mechanisms. In this manuscript we have
generated a number of higher order mutants by crossing genotypes carrying defects in each of the
short-term photoprotection mechanisms, with the final aim to obtain a direct comparison of their role
and efficiency in photoprotection. We found that mutants carrying a defect in the ΔpH-dependent
photosynthesis-control are characterized by photoinhibition of both photosystems, irrespectively of
whether PSBS-dependent NPQ or state transitions defects were present or not in the same individual,
demonstrating the primary role of PSC in photoprotection. Moreover, mutants with a limited capability
to develop a strong PSBS-dependent NPQ, were characterized by a high turnover of the D1 protein and
high values of Y(NO), which might reflect energy quenching processes occurring within the PSII reaction
center
Developing a relational meaning of the equal sign: effects of using a balance analogy in a game-based virtual environment
No full textUnderstanding the equal sign relationally is important for success in arithmetic and algebra. Yet, many elementary school children continue to perceive it operationally, as a symbol that indicates computation. Mathematics educators have long suggested the use of the balance scale as an appropriate context for developing the relational meaning of the equal sign. However, studies that directly evaluate the pedagogical effectiveness of the balance scale remain limited. This experimental study, with 148 second- and third-grade students, examined the effects of using a virtual balance scale in promoting a relational meaning of the equal sign. Findings indicate that although the virtual manipulative may be helpful in supporting a relational meaning of the equal sign, there are other features associated with the manipulative that have advantages over the use of the dynamic virtual balance scale in promoting students’ learning of mathematical equivalence. Results suggest that seemingly intuitive manipulatives that are perceptually rich may not promote optimal learning.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2018-05-01The student, Neet Priya Bajwa, accepted the attached license on 2016-04-18 at 01:49.The student, Neet Priya Bajwa, submitted this Dissertation for approval on 2016-04-18 at 02:44.This Dissertation was approved for publication on 2016-04-21 at 17:22.DSpace SAF Submission Ingestion Package generated from Vireo submission #9283 on 2016-07-07 at 13:49:41Made available in DSpace on 2016-07-07T20:27:32Z (GMT). No. of bitstreams: 3
BAJWA-DISSERTATION-2016.pdf: 20135533 bytes, checksum: 2a0152cc93406ac3bf4469f06a0db54b (MD5)
LICENSE.txt: 4213 bytes, checksum: 9b7495048df8b95c916cece3cf992e0e (MD5)
PROQUEST_LICENSE.txt: 4559 bytes, checksum: 48f9826ec20cef4c769e183c3bef6991 (MD5)
Previous issue date: 2016-04-21Embargo set by: Seth Robbins for item 93125
Lift date: 2018-07-07T20:28:14Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 93125
Lift date: 2018-07-07T20:35:34Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Only Restriction set for Item 93125 on 2018-03-27T14:24:39Z with date 2020-03-27 by [email protected] of I Only Restriction set for Item 93125 on 2018-03-27T14:25:42Z with date 2020-03-27 by [email protected] of I Only Restriction set for Item 93125 on 2018-03-27T14:25:55Z with date 2020-03-27 by [email protected] of I Only Restriction set for Item 93125 on 2018-03-27T14:25:59Z with date 2020-03-27 by [email protected] of I Only Restriction Lifted for Item 93125 on 2020-03-27T09:15:10Z.U of I Only Restriction set for Item 93125 on 2020-05-14T15:08:32Z with date 2022-05-14 by [email protected] of I Only Restriction set for Item 93125 on 2020-05-14T15:08:33Z with date 2022-05-14 by [email protected] of I Only Restriction Lifted for Item 93125 on 2022-05-14T09:15:13Z
Learning the nonlinear geometric structure of high-dimensional data: models, algorithms, and applications
No full textModern information processing relies on the axiom that high-dimensional data lie near low-dimensional geometric structures. The work presented in this thesis aims to develop new models and algorithms for learning the geometric structures underlying data and to exploit the application of geometry learning in image and video analytics. The first part of the thesis revisits the problem of data-driven learning of these geometric structures and puts forth two new nonlinear geometric models for data describing "related" objects/phenomena. The first one of these models straddles the two extremes of the subspace model and the union-of-subspaces model, and is termed the emph{metric-constrained union-of-subspaces} (MC-UoS) model. The second one of these models---suited for data drawn from a mixture of nonlinear manifolds---generalizes the kernel subspace model, and is termed the emph{metric-constrained kernel union-of-subspaces} (MC-KUoS) model. The main contributions in this regard are threefold. First, we motivate and formalize the problems of MC-UoS and MC-KUoS learning. Second, we present algorithms that efficiently learn an MC-UoS or an MC-KUoS underlying data of interest. Third, we extend these algorithms to the case when parts of the data are missing. The second part of the thesis considers the problem of learning meaningful human action attributes from video data. Representation of human actions as a sequence of human body movements or action attributes enables the development of models for human activity recognition and summarization. We first propose a hierarchical union-of-subspaces model and an approach called hierarchical sparse subspace clustering (HSSC) is developed to learn this model from the data in an unsupervised manner by capturing the variations or movements of each action in different subspaces. We then present an extension of the low-rank representation (LRR) model, termed the emph{clustering-aware structure-constrained low-rank representation} (CS-LRR) model, for unsupervised learning of human action attributes from video data. The CS-LRR model is based on the union-of-subspaces framework, and integrates spectral clustering into the LRR optimization problem for better subspace clustering results. We also introduce a hierarchical subspace clustering approach, termed hierarchical CS-LRR, to learn the attributes without the need for a priori specification of their number. By visualizing and labeling these action attributes, the hierarchical model can be used to semantically summarize long video sequences of human actions at multiple resolutions. A human action or activity can also be uniquely represented as a sequence of transitions from one action attribute to another, which can then be used for human action recognition.Ph.D.Includes bibliographical referencesby Tong W
Acquisitions and Bibliographie Services
No full textIn the unlikely event that the author did not smd UMI a complete rnanuscfïpt and there are misshg pages, these wül be m. Also, if unauthocized copyright material had to be removeâ, a note will indi the deletion. Oversize mateials (e.g., mps, drawings, chafts) am mpmduced by sectiming the original, beginning at the upper left-hand corner cvrd continuing from left to right in equal sedons wi2h maIl overlaps. Photographs induded in the original manuscn'pt have been repcoduoed xerographically in mis copy. Higher quality 6. x 9 " bladc and white photographie prints am availabie for any photographs or illustfathfts apOearing in this copy for an additional charge. Coritad UMI di- t ~ order
Some methods for statistical inference using high-dimensional linear models
No full textThe ordinary linear model has been the bedrock of signal processing, statistics, and machine learning for decades. The last decade, however, has witnessed a marked transformation of this model: instead of the classical low-dimensional setting in which the sample size exceeds the number of features/predictors/variables, we are increasingly having to operate in the high-dimensional setting in which the number of variables far exceeds the sample size. Although such high-dimensional settings would ordinarily lead to ill-posed problems, the inference task has been studied under the rubric of high-dimensional statistical inference, where various notions of structure have been imposed on the model parameters to obtain unique solutions to the inference problem. While there are many statistical methods that guarantee unique solutions, these methods can easily become computationally prohibitive in ultrahigh-dimensional settings, in which the number of variables can scale exponentially with the sample size. In other cases, the traditional notions of structure on model parameters can be rather restrictive, especially when the variables naturally appear in the form of a multi-way array (tensor), as in the case of neuroimaging data analysis.
The purpose of this dissertation is to study inference using high-dimensional linear models for the cases when (i) the number of variables can scale exponentially with the number of samples, and (ii) the variables naturally form a tensor structure. Specifically, for each of these respective cases, the dissertation (i) proposes an efficient inference approach, (ii) provides high-probability performance guarantees for the proposed approach, and (iii) demonstrates efficacy of the inference approach in statistical analysis of real-world datasets.Ph.D.Includes bibliographical reference
- …
