{"title":"Automatic word classification using simulated annealing","authors":"M. Jardino, G. Adda","doi":"10.1109/ICASSP.1993.319224","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319224","url":null,"abstract":"A bigram class model which gives the probability of a word class given its predecessor class has been developed. Simulated annealing is used to classify automatically the words of large text corpora. A first validation of the use of simulated annealing in language modeling is presented. Results are presented using a French corpus of 40000 words and a German corpus of 100000 words. It is demonstrated that simulated annealing makes it possible to classify words without any syntactic or semantic knowledge. The best results are obtained with all words gathered into a unique class at the beginning of the optimization. Simulated annealing is easy to implement and CPU time cost is not prohibitive: seven hours on a 486-33 MHz PC to classify 14000 words into 120 classes using a 75000 word training set, without any code optimization.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123875303","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}
Elisabetta Gerbino, P. Baggia, A. Ciaramella, C. Rullent
{"title":"Test and evaluation of a spoken dialogue system","authors":"Elisabetta Gerbino, P. Baggia, A. Ciaramella, C. Rullent","doi":"10.1109/ICASSP.1993.319250","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319250","url":null,"abstract":"The development of spoken dialogue systems (SDSs) requires the definition of evaluation metrics which can assess the performance of these systems at different levels and compare various SDSs. The authors present a first test, made with naive users, on an integrated dialogue system for telephone speech access to a remote data base. They describe the system architecture as well as the goals of the test, its features, the methodology used during the evaluation, and the results obtained. The SDS is shown to be effective for providing the user with the required information. The presence of spontaneous speech phenomena is frequent with naive users. The dialogue helps the user to overcome the errors due to spontaneous speech. The use of isolated words for confirmation is useful, but partially limits the interaction friendliness.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"372 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123194148","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":"Defibrillation efficacy estimation using Bayesian estimation theory","authors":"R. Malkin, T. Pilkington","doi":"10.1109/ICASSP.1993.319059","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319059","url":null,"abstract":"An implanted cardioverter/defibrillator (ICD) detects ventricular fibrillation (VF), a potentially fatal lack of electrical coordination in the heart, and then discharges a capacitor though the heart, in the hope of restoring normal rhythm. One of the clinical difficulties with implanting an ICD is efficacy estimation, estimating the correct capacitor charging voltage. Current techniques require repeated inductions of VF, and yield inaccurate and poorly defined estimates. The authors have optimized these standard efficacy estimators using Bayesian estimation theory, reducing their root-mean-square (RMS) error from 26% to 11%. They are currently developing a completely novel approach. Preliminary simulations suggest that 1 VF induction with the new approach gives efficacy-estimates with RMS errors as low as for 2 VF inductions using standard approaches.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"818 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114348911","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":"Utterance normalization using vowel features in a spoken word recognition system for multiple speakers","authors":"S. Ohno, K. Hirose, H. Fujisaki","doi":"10.1109/ICASSP.1993.319373","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319373","url":null,"abstract":"The authors propose a novel method of normalization based on linear transformation of acoustic features of input speech using only one isolated utterance each of the five vowels of Japanese by each individual speaker. Experiments on isolated word recognition combining the proposed normalization method and multiple-template DP matching showed a marked improvement in the recognition rate, especially for smaller numbers of templates per word. The proposed method gives consistently higher word recognition scores than the four-dimensional representation on the Karhunen-Loeve transformation, and also gives higher scores than the original 16-dimensional representation of filter-bank outputs, especially when the number of templates is small. Together with the fact that this method reduces the dimension of the feature vector by a factor of four, the results demonstrate the validity of the proposed method.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"81 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114003456","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":"Digital rate conversion with a non-rational ratio for high speed echo-cancellation modem","authors":"F. Ling","doi":"10.1109/ICASSP.1993.319423","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319423","url":null,"abstract":"The application of digital rate conversion in high-speed echo cancellation modems is discussed. The requirement of rate conversion in echo cancellation modems is first described. The author then introduces the concept of digital rate conversion with a nonrational ratio and considers two types of approximations for implementation of such a rate converter. Special considerations for the application of rate conversion to echo cancellation modems are discussed in detail. The design requirements and optimization criterion of interpolation filters for this application are discussed. The special problem of sampling clock jitter occurring in this application, its impact on modem performance, and its compensation techniques are discussed.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124000014","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":"Coding of interlaced sequences using a decoder-based adaptive deinterlacer","authors":"A. Zaccarin, Bede Liu","doi":"10.1109/ICASSP.1993.319826","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319826","url":null,"abstract":"Motion-compensated coding techniques developed for progressively scanned sequences are not entirely suitable for coding interlaced sequences. Direct approaches for coding interlaced sequences use the last two decoded fields for the motion-compensated prediction of the present field. The authors propose to deinterlace adaptively the last decoded field to combine in a single frame the information needed for the motion-compensated prediction. The adaptation is performed at the decoder and requires no overhead information. The proposed approach allows the use of true half-pixel accuracy motion estimates, facilitates the use of fast search algorithms when compared with the direct approaches, and has comparable performance.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124145561","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":"Instantaneous 'anything'","authors":"L. Cohen","doi":"10.1109/ICASSP.1993.319605","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319605","url":null,"abstract":"The author presents a generalization of the concepts of instantaneous frequency and group delay. A simple formula is given for obtaining the instantaneous value of any quantity. The physical interpretation which emerges from these considerations is discussed. Examples are given using time, frequency and scale.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126444613","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":"Parametric approximation of room impulse responses by multirate systems","authors":"M. Schönle, N. Fliege, U. Zölzer","doi":"10.1109/ICASSP.1993.319078","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319078","url":null,"abstract":"The authors propose an approach to the approximation and real-time simulation of room impulse responses. A time-frequency representation of impulse response data is obtained by wavelet decomposition using a multirate analysis filter bank. In a second step the parameters of cascaded moving average comb filter structures in the frequency subbands are calculated. Combining the output signals of the subband models by a synthesis filter bank with perfect reconstruction properties gives an approximation of the broadband impulse response. Due to the multirate signal decomposition with a high frequency resolution at low frequencies very long room impulse responses can be modeled accurately and simulated in real time.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126465259","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":"An architecture for the estimation of higher order cumulants","authors":"H. M. Stellakis, E. Manolakos","doi":"10.1109/ICASSP.1993.319634","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319634","url":null,"abstract":"To achieve real-time performance in signal processing applications that require the estimation of higher order statistics, it is necessary to introduce parallel processing and pipelining. The authors present a two stage VLSI architecture for the computation of all the non-negative lags of the cumulants of a real, one-dimensional data sequence. All the non-negative lags of the moments, up to the fourth order, are computed first by a triangular array, based on the indirect block-type estimation approach (C. L. Nikias and M. R. Roghuveer, 1987). The second and fourth order moments are then used to compute the fourth order cumulants using additional processors. A systematic algorithm-to-architectures synthesis methodology facilitated the design of both parts of the architecture and their optimal space and data flow matching.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126490755","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":"Real-time neural computation of the noise subspace for the MUSIC algorithm","authors":"F. Luo, Yan-Da Li","doi":"10.1109/ICASSP.1993.319161","DOIUrl":"https://doi.org/10.1109/ICASSP.1993.319161","url":null,"abstract":"A neural network approach to computing in real time the noise subspace for the MUSIC bearing estimation algorithm is proposed. The authors show analytically and by simulation results that the proposed neural network is guaranteed to provide the solution arbitrarily close to the accurate noise subspace during an elapsed time of only a few characteristic time constants of the circuit. The key features of this proposed computational approach are asynchronous parallel processing, continuous-time dynamics, and a high-speed computational compatibility.<<ETX>>","PeriodicalId":428449,"journal":{"name":"1993 IEEE International Conference on Acoustics, Speech, and Signal Processing","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1993-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125805867","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}