A comprehensive investigation of two acoustic feature sets for English stop consonants spoken in syllable initial position was conducted to determine the relative invariance of the features that cue place and voicing. The features evaluated were overall spectral shape, encoded as the cosine transform coefficients of the nonlinearly scaled amplitude spectrum, and formants. In addition, features were computed both for the static case, i.e., from one 25-ms frame starting at the burst, and for the dynamic case, i.e., as parameter trajectories over several frames of speech data. All features were evaluated with speaker-independent automatic classification experiments using the data from 15 speakers to train the classifier and the data from 15 different speakers for testing. The primary conclusions from these experiments, as measured via automatic recognition rates, are as follows: (1) spectral shape features are superior to both formants, and formants plus amplitudes; (2) features extracted from the dynamic spectrum are superior to features extracted from the static spectrum; and (3) features extracted from the speech signal beginning with the burst onset are superior to features extracted from the speech signal beginning with the vowel transition. Dynamic features extracted from the smoothed spectra over a 60-ms interval timed to begin with the burst onset appear to account for the primary vowel context effects. Automatic recognition results for the 6 stops (93.7%) based on 20 features was better than the rates obtained with human listeners for a 50-ms segment (89.9%) and only slightly worse than the rates obtained by human listeners for a 100-ms interval (96.6%). Thus the basic conclusion from our work is that dynamic spectral shape features are acoustically invariant cues for both place and voicing in initial stop consonants.

The Carbon Nanotube Field Effect Transistor (CNFET) is one of the most promising candidates to become successor of silicon CMOS in the near future because of its better electrostatics and higher mobility. The CNFET has many parameters such as operating voltage, number of tubes, pitch, nanotube diameter, dielectric constant, and contact materials which determine the digital circuit performance. This paper presents a study that investigates the effect of different CNFET parameters on performance and proposes a new CNFET design methodology to optimize performance characteristics such as current driving capability, delay, power consumption, and area for digital circuits. We investigate and conceptually explain the performance measures at 32 nm technologies for pure-CNFET, hybrid MOS-CNFET, and CMOS configurations. In our proposed design methodology, the power delay product (PDP) of the optimized CNFET is about 68%, 63%, and 79% less than that of the nonoptimized CNFET, hybrid MOS-CNFET, and CMOS circuits, respectively. Therefore, the proposed CNFET design is a strong candidate to implement high performance digital circuits.

Current insulin therapy for patients with type 1 diabetes often results in high variability in blood glucose concentration and may cause hyper-and hypoglycemic episodes. Closing the glucose control loop with a fully automated control system improves the quality of life for insulin-dependent patients. This paper presents a nonlinear model predictive control technique for glucose regulation in type 1 diabetic patients. The proposed technique uses a neural network as a nonlinear model for prediction of future glucose values and a fuzzy logic controller (FLC) to determine the insulin dose required to regulate the blood glucose level, especially after unmeasured meals. In the proposed technique, to avoid errors of meal estimation, the patient is not required to enter any data such as the meal time and size which was, in previous systems, necessary to determine the insulin bolus. The use of neural networks in predicting future glucose levels helps the proposed control strategy to handle delays associated with insulin absorption and time-lag between subcutaneous glucose readings and the plasma glucose level. The FLC uses the predicted glucose values to determine the required insulin bolus. A feed forward neural network (FFNN) and a recurrent neural network (RNN) are tested and evaluated as nonlinear glucose prediction models. Simulation results for three meal challenges are demonstrated. our results indicate that, the use of a neural network as a predictor along with a FL controller can decrease the postprandial glucose concentration, avoids hyper glycemia, and dynamically responds to glycemic challenges. The simulation results also indicate that, the use of a RNN in glucose prediction gives better results than the use of a FFNN. The RNN provides much better prediction performance than the FFNN especially at longer prediction horizons.

A novel pattern classification technique and a new feature extraction method are described and tested for vowel classification. The pattern classification technique partitions an N-way classification task into N*(N-1)/2 two-way classification tasks. Each two-way classification task is performed using a neural network classifier that is trained to discriminate the two members of one pair of categories. Multiple two-way classification decisions are then combined to form an N-way decision. Some of the advantages of the new classification approach include the partitioning of the task allowing independent feature and classifier optimization for each pair of categories, lowered sensitivity of classification performance on network parameters, a reduction in the amount of training data required, and potential for superior performance relative to a single large network. The features described in this paper, closely related to the cepstral coefficients and delta cepstra commonly used in speech analysis, are developed using a unified mathematical framework which allows arbitrary nonlinear frequency, amplitude, and time scales to compactly represent the spectral/temporal characteristics of speech. This classification approach, combined with a feature-ranking algorithm which selected the 35 most discriminative spectral/temporal features for each vowel pair, resulted in 71.5 % accuracy for classification of 16 vowels extracted from the TIMIT database. These results, significantly higher than other published results for the same task, illustrate the potential for the methods presented in this paper.

No abstract