Publications with Fulltext

Permanent URI for this collectionhttps://hdl.handle.net/20.500.14288/6

Browse

Now showing 1 - 17 of 17

Open Access
A numerical and experimental study of optimal velocity feedback control for vibration suppression of a plate-like structure
(Sage, 2015) Department of Mechanical Engineering; Department of Mechanical Engineering; Boz, Utku; Arıdoğan, Mustafa Uğur; Başdoğan, İpek; PhD Student; College of Engineering; Graduate School of Sciences and Engineering; N/A; N/A; 179940
This study presents a numerical and an experimental study on an active vibration control system. The system includes a fully-clamped plate and two surface bonded piezoelectric actuators and a collocated velocity sensor at one of the actuator locations. One of the piezoelectric actuators is used for disturbance actuation and the other one is used for control actuation. A model based optimal velocity feedback controller is used as control algorithm. The disturbance and actuator models are obtained through experimental characterization of the plate under the effect of the disturbance source. A representative SIMULINK model is built in parallel to the development of the experimental setup in order to investigate performance of the controller for various control parameters. After the model based optimal controller is designed, performance of the optimal velocity feedback controller is validated with the experimental study by comparing the vibration suppression values at multiple modes of the structure. Results show that the developed control methodology effectively suppresses the vibration amplitudes at multiple modes of the structure and also vibration attenuation levels can be predicted accurately with the simulations for various controller design parameters. It is also demonstrated that using an optimal controller enhances the performance of the system as opposed to just using velocity feedback algorithm for the active vibration control of the smart plate.
Open Access
Benefits and potential harms of COVID-19 vaccination during pregnancy: evidence summary for patient counseling
(Wiley, 2021) O'Brien, P.; Heath, P. T.; Le Doare, K.; von Dadelszen, P.; Magee, L.; Ladhani, S.; Khalil, A.; Kalafat, Erkan; Faculty Member; School of Medicine; 197389
Open Access
Continuous emotion tracking using total variability space
(International Speech Communication Association (ISCA), 2015) Department of Electrical and Electronics Engineering; Department of Computer Engineering; Department of Electrical and Electronics Engineering; Department of Computer Engineering; Khaki, Hossein; Erzin, Engin; Faculty Member; Graduate School of Sciences and Engineering; College of Engineering; N/A; 34503
Automatic continuous emotion tracking (CET) has received increased attention with expected applications in medical, robotic, and human-machine interaction areas. The speech signal carries useful clues to estimate the affective state of the speaker. In this paper, we present Total Variability Space (TVS) for CET from speech data. TVS is a widely used framework in speaker and language recognition applications. In this study, we applied TVS as an unsupervised emotional feature extraction framework. Assuming a low temporal variation in the affective space, we discretize the continuous affective state and extract i-vectors. Experimental evaluations are performed on the CreativeIT dataset and fusion results with pool of statistical functions over mel frequency cepstral coefficients (MFCCs) show a 2% improvement for the emotion tracking from speech.
Open Access
Cystic renal diseases: role of ultrasound. Part I, non-genetic cystic renal diseases
(Iuliu Hatieganu Medical Publishing House, 2021) Gündüz, Nesrin; Keven, Ayşe; Durmaz, Emel; Aslan, Mine; Aslan, Ahmet; Kabaalioğlu, Adnan; Güneyli, Serkan; Koç University Hospital
Kidney cysts are quite common in adults. Though small simple renal cysts in an adult over 30-40 years of age are not too unusual, however, if the same cysts are seen in a child, and especially if there are additional findings, then several diagnostic possibilities may come to mind. The role of ultrasound, together with the help of intravenous contrast agents and Doppler mode, is very critical in describing the morphologic features and follow-up of the complex or multiple and bilateral renal cysts. Sonographic signs are occasionally specific for diagnosis, but in many cases they should be evaluated together with the other genetic and clinical data to reach a diagnosis. The first part of this pictorial essay includes ""non-genetic cystic renal diseases"" and the second part will include ""genetic cystic renal diseases"".
Open Access
Cystic renal diseases: role of ultrasound. Part II, genetic cystic renal diseases
(Romanian Society for Ultrasonography in Medicine and Biology (SRUMB), 2022) Gündüz, Nesrin; Keven, Ayşe; Durmaz, Emel; Aslan, Mine; Aslan, Ahmet; Kabaalioğlu, Adnan; Güneyli, Serkan; Koç University Hospital
Kidney cysts are quite common in adults. Though small simple renal cysts in an adult over 30-40 years of age are not too unusual, however, if the same cysts are seen in a child, and especially if there are additional findings, then several diagnostic possibilities may come to mind. The role of ultrasound, together with the help of intravenous contrast agents and Doppler mode, are very critical in describing the morphologic features and follow-up of the complex or multiple and bilateral renal cysts. These sonographic signs are occasionally specific for diagnosis, but in many cases sonographic clues should be evalu-ated together with the other genetic and clinical data to reach diagnosis. The first part of this pictorial essay included the introduction into the subject and the classification of non-genetic cystic renal diseases. The key features for the non-genetic cystic renal diseases are illustrated. In the second part, eye-catching fea-tures of genetic cystic renal diseases are demonstrated.
Open Access
Dynamic characterization and modeling of rubber shock absorbers: a comprehensive case study
(Sage, 2018) Department of Mechanical Engineering; Department of Mechanical Engineering; Uçar, Hakan; Başdoğan, İpek; PhD Student; College of Engineering; N/A; 179940
Rubber or elastomeric materials are widely used for shock absorbers having elastic and viscous properties such as high inherent damping, deflection capacity, and energy storage. The dynamic properties of these components are of primary concern in designing rubber absorbers to reduce the shock loading given as well as the structure-borne noise transmissibility. Besides, the dynamic response of the mechanical systems, at where the rubber shock absorbers are used, is directly associated with the properties of the shock absorbers. In order to determine these properties of the rubber, mathematical models are created in terms of hyperelasticity and viscoelasticity. The hyperelastic and viscoelastic material models represent the nonlinear elastic and strain rate dependencies of the overall rubber behavior, respectively. Hyperelastic material model captures the material's nonlinear elasticity with no-time dependence whereas viscoelastic model describes the material response which contains an elastic and viscous part depending on time, frequency, and temperature. This paper presents the dynamic characterization of rubber shock absorbers, having different shore hardness values, in terms of hyperelastic and viscoelastic constitutive models. The parameters of the constitutive models are determined from the uniaxial tensile and relaxation tests. These parameters are used for the numerical model of the rubber components and the accuracy of the characterization is presented by means of a numerical case study.
Open Access
Evaluation of immunogenicity and reactogenicity of COVID-19 vaccines in pregnant women
(Wiley, 2022) Blakeway, H.; Amin Chowdhury, Z.; Prasad, S.; İsmail, M.; Abdallah, F.N.; Rezvani, A.; Amirthalingam, G.; Brown, K.; Le Doare, K.; Heath, P.T.; Ladhani, S.N.; Khalil, A.; Kalafat, Erkan; Faculty Member; School of Medicine; 197389
Objective: severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in pregnancy is associated with increased risk of adverse maternal and perinatal outcomes. Vaccines are highly effective at preventing severe coronavirus disease 2019 (COVID-19), but there are limited data on COVID-19 vaccines in pregnancy. This study aimed to investigate the reactogenicity and immunogenicity of COVID-19 vaccines in pregnant women when administered according to the 12-week-interval dosing schedule recommended in the UK. Methods: this was a cohort study of pregnant women receiving COVID-19 vaccination between April and September 2021. The outcomes were immunogenicity and reactogenicity after COVID-19 vaccination. Pregnant women were recruited by phone, e-mail and/or text and were vaccinated according to vaccine availability at their local vaccination center. For immunogenicity assessment, blood samples were taken at specific timepoints after each dose to evaluate nucleocapsid protein (N) and spike protein (S) antibody titers. The comparator group comprised non-pregnant female healthcare workers in the same age group who were vaccinated as part of the national immunization program in a contemporaneous longitudinal cohort study. Longitudinal changes in serum antibody titers and association with pregnancy status were assessed using a two-step regression approach. Reactogenicity assessment in pregnant women was undertaken using an online questionnaire. The comparator group comprised non-pregnant women aged 18–49 years who had received two vaccine doses in primary care. The association of pregnancy status with reactogenicity was assessed using logistic regression analysis. Results: overall, 67 pregnant women, of whom 66 had received a mRNA vaccine, and 79 non-pregnant women, of whom 50 had received a mRNA vaccine, were included in the immunogenicity study. Most (61.2%) pregnant women received their first vaccine dose in the third trimester, while 3.0% received it in the first trimester and 35.8% in the second trimester. SARS-CoV-2 S-antibody geometric mean concentrations after mRNA vaccination were not significantly different at 2–6 weeks after the first dose but were significantly lower at 2–6 weeks after the second dose in infection-naïve pregnant compared with non-pregnant women. In pregnant women, prior infection was associated with higher antibody levels at 2–6 weeks after the second vaccine dose. Reactogenicity analysis included 108 pregnant women and 116 non-pregnant women. After the first dose, tiredness and chills were reported less commonly in pregnant compared with non-pregnant women (P = 0.043 and P = 0.029, respectively). After the second dose, feeling generally unwell was reported less commonly (P = 0.046) in pregnant compared with non-pregnant women. Conclusions: using an extended 12-week interval between vaccine doses, antibody responses after two doses of mRNA COVID-19 vaccine were found to be lower in pregnant compared with non-pregnant women. Strong antibody responses were achieved after one dose in previously infected women, regardless of pregnancy status. Pregnant women reported fewer adverse events after both the first and second dose of vaccine. These findings should now be addressed in larger controlled studies.
Open Access
Federated dropout learning for hybrid beamforming with spatial path index modulation in multi-user MMWave-MIMO systems
(Institute of Electrical and Electronics Engineers (IEEE), 2021) Mishra, Kumar Vijay; Department of Electrical and Electronics Engineering; Department of Electrical and Electronics Engineering; Ergen, Sinem Çöleri; Elbir, Ahmet Musab; Faculty Member; College of Engineering; 7211; N/A
Millimeter wave multiple-input multiple-output (mmWave-MIMO) systems with small number of radio-frequency (RF) chains have limited multiplexing gain. Spatial path index modulation (SPIM) is helpful in improving this gain by utilizing additional signal bits modulated by the indices of spatial paths. In this paper, we introduce model-based and model-free frameworks for beamformer design in multi-user SPIM-MIMO systems. We first design the beamformers via model-based manifold optimization algorithm. Then, we leverage federated learning (FL) with dropout learning (DL) to train a learning model on the local dataset of users, who estimate the beamformers by feeding the model with their channel data. The DL randomly selects different set of model parameters during training, thereby further reducing the transmission overhead compared to conventional FL. Numerical experiments show that the proposed framework exhibits higher spectral efficiency than the state-of-the-art SPIM-MIMO methods and mmWave-MIMO, which relies on the strongest propagation path. Furthermore, the proposed FL approach provides at least 10 times lower transmission overhead than the centralized learning techniques.
Open Access
Generalized Polytopic Matrix Factorization
(Institute of Electrical and Electronics Engineers (IEEE), 2021) Department of Electrical and Electronics Engineering; Department of Electrical and Electronics Engineering; Erdoğan, Alper Tunga; Tatlı, Gökcan; Faculty Member; College of Engineering; Graduate School of Sciences and Engineering; 41624; N/A
Polytopic Matrix Factorization (PMF) is introduced as a flexible data decomposition tool with potential applications in unsupervised learning. PMF assumes a generative model where observations are lossless linear mixtures of some samples drawn from a particular polytope. Assuming that these samples are sufficiently scattered inside the polytope, a determinant maximization based criterion is used to obtain latent polytopic factors from the corresponding observations. This article aims to characterize all eligible polytopic sets that are suitable for the PMF framework. In particular, we show that any polytope whose set of vertices have only permutation and/or sign invariances qualifies for PMF framework. Such a rich set of possibilities enables elastic modeling of independent/dependent latent factors with combination of features such as relatively sparse/antisparse subvectors, mixture of signed/nonnegative components with optionally prescribed domains.
Open Access
Imaging appearance of post umbilical venous catheter displacement: liver collections in children
(Romanian Society for Ultrasonography in Medicine and Biology (SRUMB), 2021) Keven, Ayşe; Durmaz Emel; Çeken Kağan; Kabaalioğlu, Adnan; Other; Koç University Hospital
Umbilical venous catheters (UVCs) have become a part of routine perinatal care. In the case of its misplacement, extravasation into liver parenchyma might be observed and unusual findings might be detected and a suspicion of tumoral lesions emerges during the ultrasound examination. To avoid the unnecessary liver biopsies and catastrophic complications of UVC misplacement in the pediatric population, clinicians and radiologists must be familiar with the radiological findings. We aimed to present monographic and computed tomographic images of liver collections resulting from UVC malposition.
Open Access
Maternal and perinatal outcomes of SARS-CoV-2 infection in unvaccinated pregnancies during Delta and Omicron waves
(Wiley, 2022) Birol Ilter, P.; Prasad, S.; Mutlu, M.A.; Tekin, A.B.; O'Brien, P.; von Dadelszen, P.; Magee, L.A.; Tekin, S.; Tug, N.; Khalil, A.; Kalafat, Erkan; Faculty Member; School of Medicine; 197389
Objective: there is little evidence related to the effects of the Omicron severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variant on pregnancy outcomes, particularly in unvaccinated women. This study aimed to compare pregnancy outcomes of unvaccinated women infected with SARS-CoV-2 during the pre-Delta, Delta and Omicron waves. Methods: this was a retrospective cohort study conducted at two tertiary care facilities: Sancaktepe Training and Research Hospital, Istanbul, Turkey, and St George's University Hospitals NHS Foundation Trust, London, UK. Included were women who tested positive for SARS-CoV-2 by real-time reverse-transcription polymerase chain reaction (RT-PCR) during pregnancy, between 1 April 2020 and 14 February 2022. The cohort was divided into three periods according to the date of their positive RT-PCR test: (i) pre-Delta (1 April 2020 to 8 June 2021 in Turkey, and 1 April 2020 to 31 July 2021 in the UK), (ii) Delta (9 June 2021 to 27 December 2021 in Turkey, and 1 August 2021 to 27 December 2021 in the UK) and (iii) Omicron (after 27 December 2021 in both Turkey and the UK). Baseline data collected included maternal age, parity, body mass index, gestational age at diagnosis and comorbidities. The primary outcome was the need for oxygen supplementation, classified as oxygen support via nasal cannula or breather mask, non-invasive mechanical ventilation with continuous positive airway pressure (CPAP) or high-flow oxygen, mechanical ventilation with intubation, or extracorporeal membrane oxygenation (ECMO). Inferences were made after balancing of confounders, using an evolutionary search algorithm. Selected confounders were maternal age, body mass index and gestational age at diagnosis of infection. Results: during the study period, 1286 unvaccinated pregnant women with RT-PCR-proven SARS-CoV-2 infection were identified, comprising 870 cases during the pre-Delta period, 339 during the Delta wave and 77 during the Omicron wave. In the confounder-balanced cohort, infection during the Delta wave vs during the pre-Delta period was associated with increased need for nasal oxygen support (risk ratio (RR), 2.53 (95% CI, 1.75-3.65); P < 0.001), CPAP or high-flow oxygen (RR, 2.50 (95% CI, 1.37-4.56); P = 0.002), mechanical ventilation (RR, 4.20 (95% CI, 1.60-11.0); P = 0.003) and ECMO (RR, 11.0 (95% CI, 1.43-84.7); P = 0.021). The maternal mortality rate was 3.6-fold higher during the Delta wave compared to the pre-Delta period (5.3% vs 1.5%, P = 0.010). Infection during the Omicron wave was associated with a similar need for nasal oxygen support (RR, 0.62 (95% CI, 0.25-1.55); P = 0.251), CPAP or high-flow oxygen (RR, 1.07 (95% CI, 0.36-3.12); P = 0.906) and mechanical ventilation (RR, 0.44 (95% CI, 0.06-3.45); P = 0.438) with that in the pre-Delta period. The maternal mortality rate was similar during the Omicron wave and the pre-Delta period (1.3% vs 1.3%, P = 0.999). The need for nasal oxygen support during the Omicron wave was significantly lower compared to the Delta wave (RR, 0.26 (95% CI, 0.11-0.64); P = 0.003). Perinatal outcomes were available for a subset of the confounder-balanced cohort. Preterm birth before 34 weeks' gestation was significantly increased during the Delta wave compared with the pre-Delta period (15.4% vs 4.9%, P < 0.001). Conclusions: among unvaccinated pregnant women, SARS-CoV-2 infection during the Delta wave, in comparison to the pre-Delta period, was associated with increased requirement for oxygen support (including ECMO) and higher maternal mortality. Disease severity and pregnancy complications were similar between the Omicron wave and pre-Delta period. SARS-CoV-2 infection of unvaccinated pregnant women carries considerable risks of morbidity and mortality regardless of variant, and vaccination remains key. Miscommunication of the risks of Omicron infection may impact adversely vaccination uptake among pregnant women, who are at increased risk of complications related to SARS-CoV-2.
Open Access
Neonatal adverse outcome in twin pregnancies complicated by small-for-gestational age: twin vs singleton reference chart
(Wiley, 2021) Briffa, C.; Di, Fabrizio C.; Giorgione, V.; Bhate, R.; Huddy, C.; Richards, J.; Shetty, S.; Khalil, A.; Kalafat, Erkan; Faculty Member; School of Medicine; 197389
Objective: the use of twin-specific versus singleton charts in the assessment of twin pregnancies has been controversial. The aim of the study was to assess whether a diagnosis of small for gestational age (SGA) made using twin specific estimated fetal weight (EFW) and birthweight (BW) charts is more strongly associated with adverse neonatal outcomes compared to singleton charts in twin pregnancies. Methods: this was a cohort study of twin pregnancies delivered at St George's Hospital in London between January 2007 and May 2020. Twin pregnancies complicated by intrauterine demise of one or both twins; aneuploidy or major fetal abnormality, twin-to-twin transfusion syndrome or twin anemia polycythemia sequence (TAPS); and those delivered before 32 weeks' gestation, were excluded. SGA was defined as EFW or BW below the 10th centile. The main study outcome was composite neonatal morbidity, which was stratified to mild or severe for sensitivity analysis. Mixed-effects logistic regression analysis with random pregnancy level intercepts was used to test the association between different SGA classifications and adverse neonatal outcomes. Results: a total of 1329 twin pregnancies were identified, and 913 twin pregnancies (1826 infants) included in the analysis. Of these, 723 (79.2%) were dichorionic and 190 (20.8%) monochorionic. Using the singleton charts, 33.3% and 35.7% were classified as SGA by the singleton chart when using EFW and BW, respectively. The corresponding figures were 5.9% and 5.8% when using the twin specific charts. EFW SGA according to the twin charts, had a significant association with neonatal morbidity (OR 4.78, 95% CI 1.47-14.7, P=0.007), when compared to AGA twins. However, EFW below the 10th percentile according to singleton standards did not have a significant association with neonatal morbidity (OR 1.36, 95% CI 0.63-2.88, P=0.424). SGA classification of EFW using twin specific standards significantly better model fit than using singleton standard (P<0.001, likelihood ratio test). When twin charts were used for BW classification, BW SGA was significantly associated with 9.2 times increased odds of neonatal morbidity (P<0.001). Neonates classified as SGA only with singleton BW standard, but not with twin specific charts, had a significantly lower rate of adverse outcomes (OR 0.24, 95% CI 0.07-0.66, P=0.009), when compared to AGA twins. Conclusion: the singleton charts classified one third of twins as SGA, both prenatally and postnatally. SGA infants according to the twin specific charts, but not the singleton charts, had a significantly increased risk of adverse neonatal outcomes. This study provides further evidence that twin specific charts are better predictors of adverse neonatal outcomes; the use of these charts may reduce misclassification of twins as SGA and improve identification of those infants who are truly growth restricted.
Open Access
Operational response prediction for rigidly linked structures via direct inversion method
(Sage, 2016) Department of Mechanical Engineering; Department of Mechanical Engineering; Uçar, Hakan; Başdoğan, İpek; PhD Student; College of Engineering; N/A; 179940
Most of the mechanical systems are composed of different subsystems connected and coupled by several links. Any excitation acting on the system is divided into several internal forces which propagate through these links or so-called transfer paths. For these kinds of systems, it is important to predict the structure's response due to this excitation which is transmitted via propagation paths. Predicting the operational response as much as accurate at the point of interest is of great importance in terms of design optimization and condition monitoring at where the response cannot be measured due to some physical constraints. In accordance with this purpose, the identification of operational internal forces is necessary. In cases where direct measurement of the operational forces is impossible or impractical, especially for complex structures, a common approach is to identify the operational forces based on measured frequency response functions and a set of measured operational responses. The classical approach is the Moore-Penrose pseudo inversion, which needs significant number of frequency response function measurements and huge time consumption and effort since the coupled system is to be disassembled at all interfaces. Noting that, real complex structures have some physical limitations to be disassembled, more practical and faster approaches are required for real-life applications. The aim of this study is to present direct inversion method to identify the operational forces and hence, predict the operational response for rigidly linked vibrating structures and also demonstrate the effect of mass loadings of the transducers and noise during frequency response function measurements. The algorithms investigated herein are applied in numerical and experimental setups composed of rigidly linked structures.
Open Access
Ottoman royal women's spaces the acoustic dimension
(Johns Hopkins University (JHU) Press, 2008) Department of Archeology and History of Art; Department of Archeology and History of Art; Ergin, Nina Macaraig; Faculty Member; College of Social Sciences and Humanities; 52311
In their discussion of space in relation to gender, historians of women in the Middle East so far have focused primarily on physical and visual access. This paper argues that women's acoustic space merits closer consideration, especially since acoustic methods of communication very often could and did exceed the limits of vision and visually bounded space. This argument is based on three different case studies concerning Ottoman royal women of the sixteenth and seventeenth centuries: (1) harem women's auditory access to imperial council meetings; (2) common petitioners' auditory access to the mother of the sultan as she traveled by carriage through the imperial capital during her frequent processions; and (3) Qur'anic recitation and prayers as commissioned by female mosque patrons. These case studies have more wide-ranging implications in that they allow for conceptual experimentation leading towards a refinement of the categories of private/public, male/female space, based on the permeability of acoustic space.
Open Access
Synchronous overlap and add of spectra for enhancement of excitation in artificial bandwidth extension of speech
(International Speech Communication Association (ISCA), 2015) Department of Electrical and Electronics Engineering; Department of Computer Engineering; Department of Electrical and Electronics Engineering; Department of Computer Engineering; Turan, Mehmet Ali Tuğtekin; Erzin, Engin; Faculty Member; Graduate School of Sciences and Engineering; College of Engineering; N/A; 34503
In this paper, a new approach that extends narrow-band excitation signals using synchronous overlap and add (SOLA) of spectra have been proposed. Although artificial bandwidth extension (ABE) of speech has been extensively studied, the role of excitation spectra has not been as widely studied as the spectral envelope extension. In this study ABE is investigated with the widely used source-filter framework, where speech signal is decomposed into excitation signal (source) and spectral envelope (filter). For the spectral envelope extension, our former work based on hidden Markov model has been used. For the excitation signal extension, we propose a SOLA of excitation spectra, where the high end of the excitation spectra is extended by preserving the harmonic structure. In experimental studies, we also apply two other well-known extension techniques for excitation signals. Then comparatively we evaluate the overall performance of proposed system using the PESQ metric. Our findings indicate that the proposed excitation extension method delivers significant quality improvements.
Open Access
Ultrasonically-assisted synthesis of CeO(2) within WS(2) interlayers forming type II heterojunction for a VOC photocatalytic oxidation
(Elsevier, 2023) Hassandoost, R.; Yousef Tizhoosh, N.; Esmat, M.; Guselnikova, O.; Hussein N Assadi, M.; Khataee, A.; Department of Chemistry; Department of Chemistry; Doustkhah, Esmail; Researcher; Koç University Tüpraş Energy Center (KUTEM) / Koç Üniversitesi Tüpraş Enerji Merkezi (KÜTEM); College of Sciences
Here, we investigate the band structure, density of states, photocatalytic activity, and heterojunction mechanism of WS2 with CeO2 (CeO2@WS2) as a photoactive heterostructure. In this heterostructure, CeO2's growth within WS2 layers is achieved through ultrasonicating WS2 and intercalating CeO2's precursor within the WS2 interlayers, followed by hydrothermal treatment. Through a set of density functional calculations, we demonstrate that CeO2 and WS2 form an interface through a covalent bonding that can be highly stable. The electrochemical impedance spectroscopy (EIS) found that the CeO2@WS2 heterostructure exhibits a remarkably higher conductivity (22.23 mS cm(-2)) compared to either WS2 and CeO2, assignable to the interface in CeO2@WS2. Furthermore, in a physically mixed CeO2-WS2 where the interaction between particles is noncovalent, the resistance was significantly higher (0.67 mS cm(-2)), confirming that the heterostructure in the interface is covalently bonded. In addition, Mott-Schottky and the bandgap measurements through Tauc plots demonstrate that the heterojunction in CeO2 and WS2 is type II. Eventually, the CeO2@WS2 heterostructure indicated 446.7 mu mol g(-1) CO2 generation from photocatalytic oxidation of a volatile organic compound (VOC), formic acid, compared to WS2 and CeO2 alone.
Open Access
Vocal tract contour tracking in rtMRI using deep temporal regression network
(Institute of Electrical and Electronics Engineers (IEEE), 2020) Department of Electrical and Electronics Engineering; Department of Computer Engineering; Department of Electrical and Electronics Engineering; Department of Computer Engineering; Asadiabadi, Sasan; Erzin, Engin; Faculty Member; Graduate School of Sciences and Engineering; College of Engineering; N/A; 34503
Recent advances in real-time Magnetic Resonance Imaging (rtMRI) provide an invaluable tool to study speech articulation. In this paper, we present an effective deep learning approach for supervised detection and tracking of vocal tract contours in a sequence of rtMRI frames. We train a single input multiple output deep temporal regression network (DTRN) to detect the vocal tract (VT) contour and the separation boundary between different articulators. The DTRN learns the non-linear mapping from an overlapping fixed-length sequence of rtMRI frames to the corresponding articulatory movements, where a blend of the overlapping contour estimates defines the detected VT contour. The detected contour is refined at a post-processing stage using an appearance model to further improve the accuracy of VT contour detection. The proposed VT contour tracking model is trained and evaluated over the USC-TIMIT dataset. Performance evaluation is carried out using three objective assessment metrics for the separating landmark detection, contour tracking and temporal stability of the contour landmarks in comparison with three baseline approaches from the recent literature. Results indicate significant improvements with the proposed method over the state-of-the-art baselines.