MLLP-VRAIN Spanish ASR Systems for the Albayzín-RTVE 2020 Speech-To-Text Challenge: Extension

This paper describes the automatic speech recognition (ASR) systems built by the MLLP-VRAIN research group of Universitat Politècnica de València for the Albayzín-RTVE 2020 Speech-to-Text Challenge, and includes an extension of the work consisting of building and evaluating equivalent systems under the closed data conditions from the 2018 challenge. The primary system (p-streaming_1500ms_nlt) was a hybrid ASR system using streaming one-pass decoding with a context window of 1.5 seconds. This system achieved 16.0% WER on the test-2020 set. We also submitted three contrastive systems. From these, we highlight the system c2-streaming_600ms_t which, following a similar configuration as the primary system with a smaller context window of 0.6 s, scored 16.9% WER points on the same test set, with a measured empirical latency of 0.81 ± 0.09 s (mean ± stdev). That is, we obtained state-of-the-art latencies for high-quality automatic live captioning with a small WER degradation of 6% relative. As an extension, the equivalent closed-condition systems obtained 23.3% WER and 23.5% WER, respectively. When evaluated with an unconstrained language model, we obtained 19.9% WER and 20.4% WER; i.e., not far behind the top-performing systems with only 5% of the full acoustic data and with the extra ability of being streaming-capable. Indeed, all of these streaming systems could be put into production environments for automatic captioning of live media streams.

Vibration control through the robust nonlinear absorber with negative stiffness and internal resonance creation

In this study, a nonlinear absorber that works with a negative stiffness mechanism is suggested to mitigate vibration, and its effect on the reduction of vibration is investigated. The negative stiffness, which is inherently nonlinear, creates internal resonance; therefore, the vibration energy can be transmitted from low-frequency to high-frequency vibrating modes, causing vibration suppression. The nonlinear absorber is added to the primary nonlinear system, and when the main system is subjected to external resonance due to harmonic excitation, the negative stiffness parameter of absorber is so adjusted that autoparametric resonance occurs and vibration is reduced. First, the mathematical model of the system is presented and the governing differential equations of the motion are derived, and then, using the multiple scale method, the equations are solved for the case without, and with the 1:3 internal resonance. The responses and their stability are inspected, discussed, and compared. After that, the effect of negative stiffness and damping parameters on vibration amplitude reduction is investigated and the adequacy of the proposed absorber will be demonstrated by numerical analysis. Finally, the energy exchange between the primary system and the absorber will be demonstrated by plotting the responses in the state space and the displacement response Fourier spectrum.

​Genetic Polymorphism within Exon 3 of HSP90AA1 Gene and its Association Studies in Sahiwal and Crossbred Cows

Background: Heat shock proteins (HSPs) are molecular chaperones that play a critical role in recovering cells from stress and form a primary system for intra cellular self defense. They are highly conserved and play a crucial role in cellular thermo tolerance and heat stress response. Though there are many HSP genes, thermo tolerance is mainly correlated with HSP70 and HSP90 genes in Livestock species. Polymorphisms in these genes have shown an association with heat tolerance, milk production, fertility and disease susceptibility in livestock. They can be used as genetic markers for the selection of animals with better climate resilience, immune response and superior performance. Methods: The present study was carried out in Sahiwal (n=50) and Crossbred cows (n=50) with the objective to identify polymorphisms in HSP90AA1 gene. A 450 bp fragment of bovine HSP90AA1 gene covering exon3 was subjected to Polymerase Chain Reaction-Single-Strand Conformation Polymorphism (PCR-SSCP) technique to identify the polymorphism. PCR-SSCP patterns were correlated with the physiological, productive and reproductive traits in Sahiwal and crossbred cows using the univariate GLM model of SPSS 25. Result: The PCR-SSCP of exon 3 of HSP90AA1 gene yielded two conformational patterns AA and AB corresponding to two allelic variants A and B in both Sahiwal and crossbred cows. The allele frequencies of A and B were 0.78 and 0.22 and 0.84 and 0.16 in Sahiwal and crossbred cows, respectively. The association analysis of SSCP patterns revealed that genotype AA had higher lactation length in Sahiwal cows and higher total lactation milk yield and peak yield in crossbred cows.

Understanding Solution State Conformation and Aggregate Structure of Conjugated Polymers via Small Angle X-ray Scattering

Donor-acceptor (D-A) conjugated polymers are high-performance organic electronic materials that exhibit complex aggregation behavior. Understanding the solution state conformation and aggregation of conjugated polymers is crucial for controlling morphology during thin-film deposition and the subsequent electronic performance. However, a precise multiscale structure of solution state aggregates is lacking. Here, we present an in-depth small angle X-ray scattering (SAXS) analysis of the solution state structure of an isoindigo-bithiophene based D-A polymer (PII-2T) as our primary system. Modeling the system as a combination of hierarchical fibrillar aggregates mixed with dispersed polymers, we extract information about conformation and multiscale aggregation and also clarify the physical origin of features often observed but unaddressed or misinterpreted in small-angle scattering patterns of conjugated polymers. The persistence length of the D-A polymer extracted from SAXS agrees well with a theoretical model based on the dihedral potentials. Additionally, we show that the broad high q structure factor peak seen in scattering profiles can be attributed to lamellar stacking occurring within the fibril aggregates and that the low q aggregate scattering is strongly influenced by the polymer molecular weight. Overall, the SAXS profiles of D-A polymers in general exhibit a sensitive dependence on the co-existence of fibrillar aggregate and dispersed polymer chain populations. We corroborate our findings from SAXS with electron microscopy of freeze-dried samples for direct imaging of fibrillar aggregates. Finally, we demonstrate the generality of our approach by fitting the scattering profiles of a variety of D-A polymers. The results presented here establish a picture of the D-A polymer solution state structure and provide a general method of interpreting and analyzing their scattering profiles.

EXPERIMENTAL ANALYSIS OF THE VISCOUS TUNED MASS DAMPER FOR THE ATTENUATION OF STRUCTURAL RESPONSES

This paper presents a systematic experimental investigation on the performance of a Multiple Tuned Mass Dampers (MTMDs) attached to a structural system under dynamic load excitation. A Modal Experimental Analysis (EMA) of a three-story structural frame equipped with a viscous damper system was carried out through a series of shaking table tests to evaluate the performance and verify the analysis approach. Each of the TMDs consists of a mass attached to a structural floor via Thermoplastic Polyurethane (TPU) viscous bearing. Initially, the TMD was designed solely to control single mode vibration and then the mechanism is extended for the application of controlling multimode responses. The experiment demonstrated that the proposed viscous dampers exhibit good performance in reducing the response of structures under dynamic loads, and able to control both fundamental and higher vibration modes of a Multiple Degree of Freedom (MDOF) primary system effectively. It was also evident that the attachment of the air dashpot dampers to each of TMDs lead to better efficiency on controlling the amplification of the damper mass and significantly contribute to better structural modal tuning.

Model Bridge Span Traversed by a Heavy Mass: Analysis and Experimental Verification

In this work, we investigate the transient response of a model bridge traversed by a heavy mass moving with constant velocity. Two response regimes are identified, namely forced vibrations followed by free vibrations as the moving mass goes past the far support of the simply supported span of the bridge. Despite this being a classical problem in structural dynamics, there is an implicit assumption in the literature that moving loads possess masses that are at least an order of magnitude smaller than the mass of the bridge span that they traverse. This alludes to interaction problems involving secondary systems, whose presence does not alter the basic characteristics of the primary system. In our case, the dynamic properties of the bridge span during the passage of a heavy mass change continuously over time, leading to an eigenvalue problem that is time dependent. During the free vibration regime, however, the bridge recovers the expected dynamic properties corresponding to its original configuration. Therefore, the aim here is the development of a mathematical model whose numerical solution is validated by comparison with experimental results recovered from an experiment involving a scaled bridge span traversed by a rolling mass. Following that, the target is to identify regions in the transient response of the bridge span that can be used for recovering the bridge’s dynamic properties and subsequently trace the development of structural damage. In closing, the present work has ramifications in the development of structural health monitoring systems applicable to critical civil engineering infrastructure, such as railway and highway bridges.

Model Development and Transient Analysis of the WCLL BB BOP DEMO Configuration Using the Apros System Code

Extensive modelling and analytical work has been carried out considering the water-cooled lithium–lead breeding blanket (WCLL BB) balance of plant (BOP) configuration of the demonstration power plant (DEMO) using the Apros system code, developed by VTT Technical Research Centre of Finland Ltd. and Fortum. Contributing to the BOP work package of the EUROfusion Consortium, the integral plant model for dynamic analyses of the WCLL BB configuration has been updated with special attention to primary system components. Following trends of relevant neutronics modelling, a new BB model has been implemented in 2020 with the aim to obtain higher resolution output data and a more realistic thermalhydraulic feedback from the primary system. Once-through steam generator user components have been built based on CAD models conceived by BOP partners. Transient analyses have been performed providing a better picture regarding the behaviour of main components, e.g., the BB and the OTSGs, whilst highlighting possible ways to optimise the control scheme of the plant.

Shewanella oneidensis arcA Mutation Impairs Aerobic Growth Mainly by Compromising Translation

Arc (anoxic redox control), one of the most intensely investigated two-component regulatory systems in γ-proteobacteria, plays a major role in mediating the metabolic transition from aerobiosis to anaerobiosis. In Shewanella oneidensis, a research model for respiratory versatility, Arc is crucial for aerobic growth. However, how this occurs remains largely unknown. In this study, we demonstrated that the loss of the response regulator ArcA distorts the correlation between transcription and translation by inhibiting the ribosome biosynthesis. This effect largely underlies the growth defect because it concurs with the effect of chloramphenicol, which impairs translation. Reduced transcription of ArcA-dependent ribosomal protein S1 appears to have a significant impact on ribosome assembly. We further show that the lowered translation efficiency is not accountable for the envelope defect, another major defect resulting from the ArcA loss. Overall, our results suggest that although the arcA mutation impairs growth through multi-fold complex impacts in physiology, the reduced translation efficacy appears to be a major cause for the phenotype, demonstrating that Arc is a primary system that coordinates proteomic resources with metabolism in S. oneidensis.