{"title":"A TDPSOLA based concatenation technique for Bengali text to speech synthesis system Subachan","authors":"K. Swarna, Abu Naser","doi":"10.1109/ICECE.2016.7853866","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853866","url":null,"abstract":"Creating an intelligible as well as a natural text to speech synthesizer has been the ultimate goal of researchers for the past 30 years; and concatenative synthesis provides the most natural speech. It is usual to have distortions in the concatenation points in concatenative speech synthesis, and therefore generating audible clicks in the synthesized speech. To solve this problem, several signal processing concatenation algorithms exist, such as TDPSOLA, FDPSOLA, MBROLA etc. This paper addresses the problem of audible discontinuities at the concatenation points of diphones in Bengali speech synthesizer Subachan and solving it using TDPSOLA. In the process of doing this, we detected correct pitch mark locations of diphones, detected voiced and unvoiced speech frames of diphones and finally concatenated those diphones using TDPSOLA after rescaling them to remove energy mismatches. As a result, the audible clicks in the concatenation points are removed and speech with much better quality is generated.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"187 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124744577","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sharafat Ali, Tamal Sarkar, A. Day, Md. Rabiul Hasan, J. R. Mou, S. Rana, M. Islam, N. Ahmed
{"title":"Ultra-low loss THz waveguide with flat EML and near zero flat dispersion properties","authors":"Sharafat Ali, Tamal Sarkar, A. Day, Md. Rabiul Hasan, J. R. Mou, S. Rana, M. Islam, N. Ahmed","doi":"10.1109/ICECE.2016.7853869","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853869","url":null,"abstract":"This paper represents an ultra-low loss waveguide with flat effective material loss (EML) and near zero dispersion flattened properties suitable for THz wave transmission. A novel type of air hole position is proposed as the design of the photonic crystal fiber. Numerical analysis exhibits an EML of 0.562 dB/cm at the operating frequency of 1.0 THz and a flat EML of 0.57±0.0225 dB/cm from 1.0–2.0 THz band. The proposed porous core fiber shows a low EML of 0.473 dB/cm with negligible confinement and bending losses at 0.9 THz frequency. This waveguide also exhibits near zero flat dispersion of 0.7±0.3 ps/THz/cm in the frequency band from 1.2 to 2.0 THz.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"18 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120922525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. M. Mahfuz-Ur-Rahman, Md. Mazharul Islam, Md. Rabiul Islam
{"title":"Performance analysis of modulation techniques in multilevel inverters for direct grid connected photovoltaic arrays","authors":"A. M. Mahfuz-Ur-Rahman, Md. Mazharul Islam, Md. Rabiul Islam","doi":"10.1109/ICECE.2016.7853857","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853857","url":null,"abstract":"Development of renewable energy has become crucial for the global warming and shortage of fossil fuel. Solar energy is one of the promising renewable energy to meet the energy crisis. One of the big disadvantage of solar array is the efficiency level remains between 14%–25%. So maximum power extraction and reduction of associate losses is a challenge for the researchers. Multilevel inverters are being used in grid connected photovoltaic arrays to enhance the output voltage. This paper analyses the performance of different existing and proposed modulation techniques for an 11 KV system applied in 15 level H bridge inverter in Matlab environment in a view to reduce the harmonics as well as switching loss.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114747477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A hybrid deep model with HOG features for Bangla handwritten numeral classification","authors":"S. Sharif, Nabeel Mohammed, N. Mansoor, S. Momen","doi":"10.1109/ICECE.2016.7853957","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853957","url":null,"abstract":"Considering the practical significances, handwriting recognition is getting an intense interest to the research community. Through, several studies have been conducted for Bengali handwriting recognition, a robust model for Bengali numerals classification is still due. Therefore, a hybrid model is presented in this paper, which aims to classify the Bengali numerals more precisely. The proposed model bridges hand crafted feature extraction based approaches with the automatically learnt features of Convolutional Neural networks (CNN). It is observed that the proposed model outperforms existing models with lesser epochs. The proposed model is trained and tested with the ISI numeral dataset and also cross-validated with the CAMTERDB numeral dataset. For both scenarios, proposed model shows consistency and demonstrate the maximum accuracy of 99.02% and 99.17%, respectively. For the CMATERDB collection, the proposed model achieves the best accuracy rate reported till date.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126373099","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Traffic modelling of WiMAX under BMAP with length of packet of general distribution","authors":"Jesmin Akhter, M. Islam, M. Amin","doi":"10.1109/ICECE.2016.7853860","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853860","url":null,"abstract":"The relation between offered traffic and the length of frame taking the batch size as a parameter are nonlinearly related under Batch Markovian Arrival Process (BMAP). Therefore the prediction of carried traffic and blocking probability against the length of frames or the batched size becomes more complicated compared to conventional packet or circuit switched traffic. The previous literature deals with M/1/G/K traffic for a network in generalized form but in this paper we bring the traffic model for WiMAX network. We compare the performance of subscriber station (SS) WiMAX network for M/M/1/K and M/1/G/K traffic cases considering BMAP. Finally we show the interrelationship among: offered traffic, the length of frame, carried traffic(related to throughput), the length of queue, the size of batch and blocking probability (related to packet loss) both analytically and graphically; where the previous literature show the results partially.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126442640","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Sensitivity of a nanodoped polymer based X-ray image detector","authors":"Naimul Hassan, S. A. Mahmood","doi":"10.1109/ICECE.2016.7853879","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853879","url":null,"abstract":"The X-ray sensitivity of direct conversion X-ray image detectors based on poly(triarylamine) (PTAA) doped with bismuth oxide (Bi2O3) nanoparticle is determined for mammography. Using COMSOL Multiphysics 5.0, the continuity equations are solved considering drift of charge carriers in the presence of deep traps. The recombination between drifting charge carriers and nonuniform electric field across the photoconductor layer is also considered. Poole-Frenkel mechanism has been considered for electric field dependent mobility of charge carriers. A very promising X-ray sensitivity of 1.3295µC/cm2R has been found at an applied electric field of 10V/µm. The calculated sensitivity has been compared with that of the commercially available amorphous selenium (a-Se) based direct conversion X-ray image detector.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128007935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"‘Δ’ slotted compact printable orientation insensitive chipless RFID tag for long range applications","authors":"Md. Aminul Islam, Yixian Yap, N. Karmakar","doi":"10.1109/ICECE.2016.7853911","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853911","url":null,"abstract":"A novel design concept of a compact printable chipless RFID tag is presented. The tag consists of a triangular patch loaded with multiple triangular ‘Δ’ shaped slot resonators. Six copies of the identical patches are placed together to achieve a very compact long range chipless RFID tag. This symmetric frequency domain based tag has the advantage to be read from any orientation with the reader antennas. The tag can be read by chipless RFID tag reader in both near field and far-field measurements. This tag does not have a ground plane and has higher data density compared to most of the existing printable chipless tags.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122243785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Rahman, Md. Nur-Al-Safa Bhuiyan, Muhammad Sajjadur Rahim, Sabbir Ahmed
{"title":"A lightweight PAPR reduction scheme using Greatest Common Divisor matrix based SLM technique","authors":"M. Rahman, Md. Nur-Al-Safa Bhuiyan, Muhammad Sajjadur Rahim, Sabbir Ahmed","doi":"10.1109/ICECE.2016.7853964","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853964","url":null,"abstract":"High Peak to Average Power Ratio (PAPR) is one of the major drawbacks for Orthogonal Frequency Division Multiplexing (OFDM) transmitted signal. Among the PAPR reduction techniques, Selected Mapping (SLM) technique is one of the promising one but results in added computational complexity. In this paper, we applied Greatest Common Divisor (GCD) sequence based SLM method using our proposed reverse searching technique. The proposed scheme is named lightweight as it is very simple to implement and offers improved performance through significant reduction in PAPR and computational complexity. Additionally, we explored side information free transmission with GCD default index that achieves higher throughput but sacrifices slight PAPR reduction.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134020058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Ferdous, P. García, Mohammad Abdul Moin Oninda, M. Hoque
{"title":"MTPA and Field Weakening Control of Synchronous Reluctance motor","authors":"S. Ferdous, P. García, Mohammad Abdul Moin Oninda, M. Hoque","doi":"10.1109/ICECE.2016.7853991","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853991","url":null,"abstract":"An improved algorithm for MTPA (Maximum Torque Per Ampere) and Field Weakening operation which is suitable for implementation with Field Oriented Control, has been proposed for the Synchronous Reluctance motor (SynRel) for high dynamic performance. The proposed method overcomes the problem of excessive current flow into the machine while generating the d-axis and q-axis current commands independently. At the same time to operate the inverter in the linear modulation region (0≤m≤1) to minimize the output torque ripple. The proposed method helps to attain MTPA operation below the base speed to operate with the rated load and a fast speed response in Field Weakening mode. The validity and efficacy of the proposed algorithm is verified by simulation in MATLAB and SIMULINK.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"172 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133965480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A deep domain adaption approach for object recognition using Multiple Model Consistency analysis","authors":"B. Pal, Boshir Ahmed","doi":"10.1109/ICECE.2016.7853982","DOIUrl":"https://doi.org/10.1109/ICECE.2016.7853982","url":null,"abstract":"Domain adaption tends to transfer knowledge across domains following dissimilar distribution and where target domain has inadequate labelled samples. When knowledge is transferred from abundantly irrelevant sources negative transfer may occur resulting in poor classification of test samples. Deep learning research illustrates the semantic clustering as well as transferability of deep convolutional features for numerous tasks including domain adaption. Traditional clustering based domain adaption approaches are practical to handle negative transfer scenario. This paper presents a scheme that uses graph based consistency analysis of one supervised and another unsupervised model to effectively transfer knowledge using deep features. This approach uses local neighbourhood analysis to classify hard samples that are identified using consistency analysis of models. This method yields encouraging experimental results on benchmark domain adaption dataset compared to a single deep feature based supervised support vector machine classifier, demonstrating effective use of target domain data.","PeriodicalId":122930,"journal":{"name":"2016 9th International Conference on Electrical and Computer Engineering (ICECE)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133456080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}