Existing methods on video-based action recognition are generally view-dependent, i.e., performing recognition from the same views seen in the training data. We present a novel multiview spatio-temporal AND-OR graph (MST-AOG) representation for cross-view action recognition, i.e., the recognition is performed on the video from an unknown and unseen view. As a compositional model, MST-AOG compactly represents the hierarchical combinatorial structures of cross-view actions by explicitly modeling the geometry, appearance and motion variations. This paper proposes effective methods to learn the structure and parameters of MST-AOG. The inference based on MST-AOG enables action recognition from novel views. The training of MST-AOG takes advantage of the 3D human skeleton data obtained from Kinect cameras to avoid annotating enormous multi-view video frames, which is error-prone and time-consuming, but the recognition does not need 3D information and is based on 2D video input. A new Multiview Action3D dataset has been created and will be released. Extensive experiments have demonstrated that this new action representation significantly improves the accuracy and robustness for cross-view action recognition on 2D videos.

No abstract

No abstract

Summary Model organisms and human studies have led to increasing empirical evidence that interactions among genes contribute broadly to genetic variation of complex traits. In the presence of gene-by-gene interactions, the dimensionality of the feature space becomes extremely high relative to the sample size. This imposes a significant methodological challenge in identifying gene-by-gene interactions. In the present paper, through a Gaussian graphical model framework, we translate the problem of identifying gene-by-gene interactions associated with a binary trait D into an inference problem on the difference of two high-dimensional precision matrices, which summarize the conditional dependence network structures of the genes. We propose a procedure for testing the differential network globally that is particularly powerful against sparse alternatives. In addition, a multiple testing procedure with false discovery rate control is developed to infer the specific structure of the differential network. Theoretical justification is provided to ensure the validity of the proposed tests and optimality results are derived under sparsity assumptions. A simulation study demonstrates that the proposed tests maintain the desired error rates under the null and have good power under the alternative. The methods are applied to a breast cancer gene expression study.

Abstract. Return-Oriented Programming (ROP) is a new technique that helps the attacker construct malicious code mounted on x86/SPARC executables without any function call at all. Such technique makes the ROP malicious code contain no instruction, which is different from existing attacks. Moreover, it hides the malicious code in benign code. Thus, it circumvents the approaches that prevent control flow diversion outside legitimate regions (such as W ⊕ X) and most malicious code scanning techniques (such as anti-virus scanners). However, ROP has its own intrinsic feature which is different from normal program design: (1) uses short instruction sequence ending in "ret", which is called gadget, and (2) executes the gadgets contiguously in specific memory space, such as standard GNU libc. Based on the features of the ROP malicious code, in this paper, we present a tool DROP, which is focused on dynamically detecting ROP malicious code. Preliminary experimental results show that DROP can efficiently detect ROP malicious code, and have no false positives and negatives.

For a hydrogel coating on a substrate to be stable, covalent bonds polymerize monomer units into polymer chains, crosslink the polymer chains into a polymer network, and interlink the polymer network to the substrate. In existing methods of hydrogel coating, the three processespolymerization, crosslinking, and interlinking-usually concur. This concurrency is unnecessary and hinders the widespread applications. In particular, many hydrogels are made by free-radical polymerization, involving toxic monomers, toxic initiators, and oxygen-free environment. For example, in the free-radical polymerization of a covalently crosslinked polyacrylamide hydrogel, when subject to UV light, the vinyl groups of acrylamide monomer and N,N′-methylenebisacrylamide crosslinker are activated concurrently. The former results in polyacrylamide chains and the latter results in a polyacrylamide network. When the substrate is involved, polymerization of monomers, crosslinking of poly mer chains, and interlinking between the hydrogel and substrate proceed concurrently. Since free radical polymerization is sensitive to oxygen, molding is usually required. Besides, molding is also used to control the shape and thickness of hydrogel coating. Depending on the geometry of the substrate, coating can be technically challenging for highly curved surfaces (e.g., 1D structures), or even impossible for hollow or cage structures. Furthermore, the monomers (e.g., acrylamide, acrylic acid, etc.) are usually toxic. Consequently, free-radical polymerization is unsuitable for everyday operation.For a hydrogel coating on a substrate to be stable, covalent bonds poly merize monomer units into polymer chains, crosslink the polymer chains into a polymer network, and interlink the polymer network to the substrate. The three processes-polymerization, crosslinking, and interlinking-usually concur. This concurrency hinders widespread applications of hydrogel coatings. Here a principle is described to create hydrogel paints that decouple polymerization from crosslinking and interlinking. Like a common paint, a hydrogel paint divides the labor between the paint maker and the paint user. The paint maker formulates the hydrogel paint by copolymerizing monomer units and coupling agents into polymer chains, but does not crosslink them. The paint user applies the paint on various materials (elastomer, plastic, glass, ceramic, or metal), and by various operations (brush, cast, dip, spin, or spray). During cure, the coupling agents crosslink the polymer chains into a network and interlink the polymer network to the substrate. As an example, hydrogels with thickness in the range of 2-20 µm are dip coated on medical nitinol wires. The coated wires reduce friction by eightfold, and remain stable over 50 test cycles. Also demonstrated are several proofofconcept applica tions, including stimuliresponsive structures and antifouling model boats.A hydrogel-coated substrate unites the superior properties of the substrate (e.g., strength, stiffness, and toughness) and the superior p...

No abstract

Background/Aims: Endothelial-to-mesenchymal transition (EndMT) plays significant roles under various pathological conditions including cardiovascular diseases, fibrosis, and cancer. EndMT of endothelial progenitor cells (EPCs) contributes to neointimal hyperplasia following cell therapy Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) is a long non-coding RNA (lncRNA) that promotes metastasis and cancer. MicroRNA-145 (miR-145) is a tumor suppressor that has been reported to inhibit SMAD3-mediated epithelial-to-mesenchymal transition (EMT) of cancer cells. In the present study, we investigated the role of MALAT1 and miR-145 in EndMT of human circulating EPCs induced by transforming growth factor beta1 (TGF-β1). Methods: Human circulating EPCs were isolated and characterized by fluorescence-activated cell sorting (FACS). Expression levels of EndMT markers were assessed by qRT-PCR and western blotting. Alpha-smooth muscle actin (α-SMA) expression was measured by cell immunofluorescence staining. The regulatory relationship between MALAT1 and miR-145 and its target genes, TGFBR2 (TGFβ receptortype II) and SMAD3 (mothers against decapentaplegic homolog 3) was analyzed using the luciferase reporter assay. Results: We found that EndMT of EPCs induced by TGF-β1 is accompanied by increased MALAT1 expression and decreased miR-145 expression, and MALAT1 and miR-145 directly bind and reciprocally repress each other in these cells. Dual-Luciferase Reporter assay indicated that miR-145 inhibits TGF-β1-induced EndMT by directly targeting TGFBR2 and SMAD3. Conclusions: MALAT1 modulates TGF-β1-induced EndMT of EPCs through regulation of TGFBR2 and SMAD3 via miR-145. Thus, the MALAT1-miR-145-TGFBR2/SMAD3 signaling pathway plays a key role in TGF-β1-induced EndMT.