No abstract

Pedestrian detection is a key problem in computer vision and is currently addressed with increasingly complex solutions involving compute-intensive features and classification schemes. In this scope, Histogram of Oriented Gradients (HOG) in conjunction with linear SVM classifier is considered to be the single most discriminative feature which has been adopted as a stand-alone detector as well as a key instrument in advance systems involving hybrid features and cascaded detectors. In this paper, we propose a pedestrian detection framework which is computationally less expensive as well as more accurate than HOG-linear SVM. The proposed scheme exploits the discriminating power of the locally significant gradients in building orientation histograms without involving complex floating point operations while computing the feature. The integeronly feature allows the use of powerful Histogram Intersection Kernel SVM classifier in a fast look-up table based implementation. Resultantly, the proposed framework achieves at least 3% more accurate detection results than HOG on standard datasets while being 1.8 and 2.6 times faster on conventional desktop PC and embedded ARM platforms respectively for a single scale pedestrian detection on VGA resolution video. Additionally, hardware implementation on Altera Cyclone IV FPGA results in more than 40% savings in logic resources compared to its HOG-Linear SVM competitor. Hence, the proposed feature and classification setup is shown to be a better candidate as the single most discriminative pedestrian detector than currently accepted HOG-linear SVM.

Surface-enhanced Raman spectroscopy (SERS)-based protein analysis is a promising alternative to existing early stage diagnoses. However, SERS research conducted thus far accompanies challenges such as nonuniformity of plasmonic nanostructures, irregular coating of analytes, and denaturation of proteins, which seriously limit the practicability of suggested approaches. Here, we introduce a carboxylic acid-functionalized and graphitic nanolayer-coated three-dimensional SERS substrate (CGSS) fabricated by sequential nanotransfer printing. The substrate consists of well-defined, uniform gold nanowire arrays for effective Raman signal enhancement and a strong protein-immobilization layer. With an enhancement factor (EF) of 5.5 × 105, on par with the highest ever reported values, the CGSS allows the detection of protein conformational changes and the determination of protein concentration via Raman measurements. Exploiting the CGSS, we successfully measured the SERS spectra of Alzheimer’s biomarkers, tau protein and amyloid β, based on which secondary structural changes were analyzed quantitatively.

Application specific instructions play an important role in reducins the required code size and increasing pe@orm-ante. This paper describes a ne)v approach to generate application specific instructions for DSP applications. The proposed approach is based on a modl~ed subset-sum problem, and can support multi-cycle complex instructions as Jvell as sinsle cycle instructions,~vhile the previous state-of-the-art approaches can Senerate only the singlecycle instructions or can just select instructions from the @ed super-set of possible instructions. In addition, the proposed approach can also be applicable to the case that instructions are predefine. The expen.mental results on real applications sho~v that the proposed approach is effective in making the instructions meet the given constraints )+~ithout attachins special hard~vareaccelerators.

No abstract

As the CMOS technology enters the deep submicron design era, the lateral inter-wire coupling capacitance becomes the dominant part of load capacitance and makes RC delay on the bus structures very data-dependent.Reducing the crosscoupling capacitance is crucial for achieving high-speed as well as lower power operation. In this paper, we propose two interconnect layout design methodologies for minimizing the "cross-coupling effect' in the design of full-custom datapath. Firstly, we describe the control signal ordering scheme which was shown to minimize the switching power consumption by 10% and wire delay by 15% for a given set of benchmark examples. Secondly, a track assignment algorithm based on evolutionary programming was used to minimize the crosscoupling capacitance. Experimental results have shown that the chip performance improvement as much as 40% can be obtained using the proposed interconnect schemes in various stages of the datapath layout optimization.

We have studied the role of defects in electrolyte-gated graphene mesh (GM) field-effect transistors (FETs) by introducing engineered edge defects in graphene (Gr) channels. Compared with Gr-FETs, GM-FETs were characterized as having large increments of Dirac point shift (∼30-100 mV/pH) that even sometimes exceeded the Nernst limit (59 mV/pH) by means of electrostatic gating of H(+) ions. This feature was attributed to the defect-mediated chemisorptions of H(+) ions to the graphene edge, as supported by Raman measurements and observed cycling characteristics of the GM FETs. Although the H(+) ion binding to the defects increased the device response to pH change, this binding was found to be irreversible. However, the irreversible component showed relatively fast decay, almost disappearing after 5 cycles of exposure to solutions of decreasing pH value from 8.25 to 6.55. Similar behavior could be found in the Gr-FET, but the irreversible component of the response was much smaller. Finally, after complete passivation of the defects, both Gr-FETs and GM-FETs exhibited only reversible response to pH change, with similar magnitude in the range of 6-8 mV/pH.

Cyanuric acid hydrolase (CAH) catalyzes the hydrolytic ring-opening of cyanuric acid (2,4,6-trihydroxy-1,3,5-triazine), an intermediate in s-triazine bacterial degradation and a by-product from disinfection with trichloroisocyanuric acid. In the present study, an X-ray crystal structure of the CAH-barbituric acid inhibitor complex from Azorhizobium caulinodans ORS 571 has been determined at 2.7 Å resolution. The CAH protein fold consists of three structurally homologous domains forming a β-barrel-like structure with external α-helices that result in a three-fold symmetry, a dominant feature of the structure and active site that mirrors the three-fold symmetrical shape of the substrate cyanuric acid. The active site structure of CAH is similar to that of the recently determined AtzD with three pairs of active site Ser-Lys dyads. In order to determine the role of each Ser-Lys dyad in catalysis, a mutational study using a highly sensitive, enzyme-coupled assay was conducted. The 109-fold loss of activity by the S226A mutant was at least ten times lower than that of the S79A and S333A mutants. In addition, bioinformatics analysis revealed the Ser226/Lys156 dyad as the only absolutely conserved dyad in the CAH/barbiturase family. These data suggest that Lys156 activates the Ser226 nucleophile which can then attack the substrate carbonyl. Our combination of structural, mutational, and bioinformatics analyses differentiates this study and provides experimental data for mechanistic insights into this unique protein family.