A coupled discriminative dictionary and transformation learning approach with applications to cross domain matching

Cross domain and cross-modal matching has many applications in the field of computer vision and pattern recognition. A few examples are heterogeneous face recognition, cross view action recognition, etc. This is a very challenging task since the data in two domains can differ significantly. In this work, we propose a coupled dictionary and transformation learning approach that models the relationship between the data in both domains. The approach learns a pair of transformation matrices that map the data in the two domains in such a manner that they share common sparse representations with respect to their own dictionaries in the transformed space. The dictionaries for the two domains are learnt in a coupled manner with an additional discriminative term to ensure improved recognition performance. The dictionaries and the transformation matrices are jointly updated in an iterative manner. The applicability of the proposed approach is illustrated by evaluating its performance on different challenging tasks: face recognition across pose, illumination and resolution, heterogeneous face recognition and cross view action recognition. Extensive experiments on five datasets namely, CMU-PIE, Multi-PIE, ChokePoint, HFB and IXMAS datasets and comparisons with several state-of-the-art approaches show the effectiveness of the proposed approach. (C) 2015 Elsevier B.V. All rights reserved.

Pd-Catalyzed Hydroborylation of Alkynes: A Ligand Controlled Regioselectivity Switch for the Synthesis of alpha- or beta-Vinylboronates

A ligand controlled selective hydroborylation of alkynes to alpha- or beta-vinylboronates has been developed using a Pd catalyst. The high alpha-selectivity displayed by this reaction can be switched to furnish beta-vinylboronates by altering the ligand from a trialkylphosphine to N-heterocyclic carbene. A variety of terminal alkynes are shown to furnish the corresponding alpha- or beta-vinylboronates in good to excellent selectivity and yield. The mechanistic studies suggest that the solvent is the proton source and bromobenzene functions as an important additive in driving this reaction forward.

A coupled discriminative dictionary and transformation learning approach with applications to cross domain matching

Cross domain and cross-modal matching has many applications in the field of computer vision and pattern recognition. A few examples are heterogeneous face recognition, cross view action recognition, etc. This is a very challenging task since the data in two domains can differ significantly. In this work, we propose a coupled dictionary and transformation learning approach that models the relationship between the data in both domains. The approach learns a pair of transformation matrices that map the data in the two domains in such a manner that they share common sparse representations with respect to their own dictionaries in the transformed space. The dictionaries for the two domains are learnt in a coupled manner with an additional discriminative term to ensure improved recognition performance. The dictionaries and the transformation matrices are jointly updated in an iterative manner. The applicability of the proposed approach is illustrated by evaluating its performance on different challenging tasks: face recognition across pose, illumination and resolution, heterogeneous face recognition and cross view action recognition. Extensive experiments on five datasets namely, CMU-PIE, Multi-PIE, ChokePoint, HFB and IXMAS datasets and comparisons with several state-of-the-art approaches show the effectiveness of the proposed approach. (C) 2015 Elsevier B.V. All rights reserved.

Anomalous reduction in domain wall displacement at the morphotropic phase boundary of the piezoelectric alloy system PbTiO3-BiScO3

A comparative study of field-induced domain switching and lattice strain was carried out by in situ electric-field-dependent high-energy synchrotron x-ray diffraction on a morphotropic phase boundary (MPB) and a near-MPB rhombohedral/pseudomonoclinic composition of a high-performance piezoelectric alloy (1-x) PbTiO3-(x)BiScO3. It is demonstrated that the MPB composition showing large d(33) similar to 425 pC/N exhibits significantly reduced propensity of field-induced domain switching as compared to the non-MPB rhombohedral composition (d(33) similar to 260 pC/N). These experimental observations contradict the basic premise of the martensitic-theory-based explanation which emphasizes on enhanced domain wall motion as the primary factor for the anomalous piezoelectric response in MPB piezoelectrics. Our results favor field-induced structural transformation to be the primary mechanism contributing to the large piezoresponse of the critical MPB composition of this system.

Electrical and Plasmonic Properties of Ligand-Free Sn4+-Doped In2O3 (ITO) Nanocrystals

Sn4+-doped In2O3 (ITO) is a benchmark transparent conducting oxide material. We prepared ligand-free but colloidal ITO (8nm, 10% Sn4+) nanocrystals (NCs) by using a post-synthesis surface-modification reaction. (CH3)(3)OBF4 removes the native oleylamine ligand from NC surfaces to give ligand-free, positively charged NCs that form a colloidal dispersion in polar solvents. Both oleylamine-capped and ligand-free ITO NCs exhibit intense absorption peaks, due to localized surface plasmon resonance (LSPR) at around =1950nm. Compared with oleylamine-capped NCs, the electrical resistivity of ligand-free ITO NCs is lower by an order of magnitude (approximate to 35mcm(-1)). Resistivity over a wide range of temperatures can be consistently described as a composite of metallic ITO grains embedded in an insulating matrix by using a simple equivalent circuit, which provides an insight into the conduction mechanism in these systems.

New insight into the architecture of oxy-anion pocket in unliganded conformation of GAT domains: A MD-simulation study

Human Guanine Monophosphate Synthetase (hGMPS) converts XMP to GMP, and acts as a bifunctional enzyme with N-terminal ``glutaminase'' (GAT) and C-terminal ``synthetase'' domain. The enzyme is identified as a potential target for anticancer and immunosuppressive therapies. GAT domain of enzyme plays central role in metabolism, and contains conserved catalytic residues Cys104, His190, and Glu192. MD simulation studies on GAT domain suggest that position of oxyanion in unliganded conformation is occupied by one conserved water molecule (W1), which also stabilizes that pocket. This position is occupied by a negatively charged atom of the substrate or ligand in ligand bound crystal structures. In fact, MD simulation study of Ser75 to Val indicates that W1 conserved water molecule is stabilized by Ser75, while Thr152, and His190 also act as anchor residues to maintain appropriate architecture of oxyanion pocket through water mediated H-bond interactions. Possibly, four conserved water molecules stabilize oxyanion hole in unliganded state, but they vacate these positions when the enzyme (hGMPS)-substrate complex is formed. Thus this study not only reveals functionally important role of conserved water molecules in GAT domain, but also highlights essential role of other non-catalytic residues such as Ser75 and Thr152 in this enzymatic domain. The results from this computational study could be of interest to experimental community and provide a testable hypothesis for experimental validation. Conserved sites of water molecules near and at oxyanion hole highlight structural importance of water molecules and suggest a rethink of the conventional definition of chemical geometry of inhibitor binding site.

Mycobacteria-responsive sonic hedgehog signaling mediates programmed death-ligand 1-and prostaglandin E-2-induced regulatory T cell expansion

CD4(+)CD25(+)FoxP3(+) regulatory T cells (Tregs) are exploited by mycobacteria to subvert the protective host immune responses. The Treg expansion in the periphery requires signaling by professional antigen presenting cells and in particularly dendritic cells (DC). However, precise molecular mechanisms by which mycobacteria instruct Treg expansion via DCs are not established. Here we demonstrate that mycobacteria-responsive sonic hedgehog (SHH) signaling in human DCs leads to programmed death ligand-1 (PD-L1) expression and cyclooxygenase (COX)-2-catalyzed prostaglandin E-2 (PGE(2)) that orchestrate mycobacterial infection-induced expansion of Tregs. While SHH-responsive transcription factor GLI1 directly arbitrated COX-2 transcription, specific microRNAs, miR-324-5p and miR-338-5p, which target PD-L1 were downregulated by SHH signaling. Further, counter-regulatory roles of SHH and NOTCH1 signaling during mycobacterial-infection of human DCs was also evident. Together, our results establish that Mycobacterium directs a fine-balance of host signaling pathways and molecular regulators in human DCs to expand Tregs that favour immune evasion of the pathogen.

CROWD FLOW SEGMENTATION IN COMPRESSED DOMAIN USING CRF

Crowd flow segmentation is an important step in many video surveillance tasks. In this work, we propose an algorithm for segmenting flows in H.264 compressed videos in a completely unsupervised manner. Our algorithm works on motion vectors which can be obtained by partially decoding the compressed video without extracting any additional features. Our approach is based on modelling the motion vector field as a Conditional Random Field (CRF) and obtaining oriented motion segments by finding the optimal labelling which minimises the global energy of CRF. These oriented motion segments are recursively merged based on gradient across their boundaries to obtain the final flow segments. This work in compressed domain can be easily extended to pixel domain by substituting motion vectors with motion based features like optical flow. The proposed algorithm is experimentally evaluated on a standard crowd flow dataset and its superior performance in both accuracy and computational time are demonstrated through quantitative results.

ON THE UTILITY OF CANONICAL CORRELATION ANALYSIS FOR DOMAIN ADAPTATION IN MULTI-VIEW HEADPOSE ESTIMATION

The utility of canonical correlation analysis (CCA) for domain adaptation (DA) in the context of multi-view head pose estimation is examined in this work. We consider the three problems studied in 1], where different DA approaches are explored to transfer head pose-related knowledge from an extensively labeled source dataset to a sparsely labeled target set, whose attributes are vastly different from the source. CCA is found to benefit DA for all the three problems, and the use of a covariance profile-based diagonality score (DS) also improves classification performance with respect to a nearest neighbor (NN) classifier.