Abstract: ABET accredited programs must assess the development of student learning outcomes (SOs) in engineering programs. Different institutions implement different strategies for this assessment, and they are usually designed “in house.” This paper presents a proposal for including standardized tests to complement the ABET assessment model in an engineering college made up of six distinct engineering programs. The engineering college formulated a model of quality assurance in education to be implemented throughout the six engineering programs to regularly assess and evaluate the achievement of SOs in each program offered. The model uses diverse techniques and sources of data to assess student performance and to implement actions of improvement based on the results of this assessment. The model is called “Assessment Process Model” and it includes SOs A through K, as defined by ABET. SOs can be divided into two categories: “hard skills” and “professional skills” (soft skills). The first includes abilities, such as: applying knowledge of mathematics, science, and engineering and designing and conducting experiments, as well as analyzing and interpreting data. The second category, “professional skills”, includes communicating effectively, and understanding professional and ethnical responsibility. Within the Assessment Process Model, various tools were used to assess SOs, related to both “hard” as well as “soft” skills. The assessment tools designed included: rubrics, surveys, questionnaires, and portfolios. In addition to these instruments, the Engineering College decided to use tools that systematically gather consistent quantitative data. For this reason, an in-house exam was designed and implemented, based on the curriculum of each program. Even though this exam was administered during various academic periods, it is not currently considered standardized. In 2017, the Engineering College included three standardized tests: one to assess mathematical and scientific reasoning and two more to assess reading and writing abilities. With these exams, the college hopes to obtain complementary information that can help better measure the development of both hard and soft skills of students in the different engineering programs. In the first semester of 2017, the three exams were given to three sample groups of students from the six different engineering programs. Students in the sample groups were either from the first, fifth, and tenth semester cohorts. At the time of submission of this paper, the engineering college has descriptive statistical data and is working with various statisticians to have a more in-depth and detailed analysis of the sample group of students’ achievement on the three exams. The overall objective of including standardized exams in the assessment model is to identify more precisely the least developed SOs in order to define and implement educational strategies necessary for students to achieve them in each engineering program.
Abstract: The process in which the complementary information from multiple images is integrated to provide composite image that contains more information than the original input images is called image fusion. Medical image fusion provides useful information from multimodality medical images that provides additional information to the doctor for diagnosis of diseases in a better way. This paper represents the wavelet based medical image fusion algorithm on different multimodality medical images. In order to fuse the medical images, images are decomposed using Redundant Wavelet Transform (RWT). The high frequency coefficients are convolved with morphological operator followed by the maximum-selection (MS) rule. The low frequency coefficients are processed by MS rule. The reconstructed image is obtained by inverse RWT. The quantitative measures which includes Mean, Standard Deviation, Average Gradient, Spatial frequency, Edge based Similarity Measures are considered for evaluating the fused images. The performance of this proposed method is compared with Pixel averaging, PCA, and DWT fusion methods. When compared with conventional methods, the proposed framework provides better performance for analysis of multimodality medical images.
Abstract: A state of the art Speaker Identification (SI) system
requires a robust feature extraction unit followed by a speaker
modeling scheme for generalized representation of these features.
Over the years, Mel-Frequency Cepstral Coefficients (MFCC)
modeled on the human auditory system has been used as a standard
acoustic feature set for speech related applications. On a recent
contribution by authors, it has been shown that the Inverted Mel-
Frequency Cepstral Coefficients (IMFCC) is useful feature set for
SI, which contains complementary information present in high
frequency region. This paper introduces the Gaussian shaped filter
(GF) while calculating MFCC and IMFCC in place of typical
triangular shaped bins. The objective is to introduce a higher
amount of correlation between subband outputs. The performances
of both MFCC & IMFCC improve with GF over conventional
triangular filter (TF) based implementation, individually as well as
in combination. With GMM as speaker modeling paradigm, the
performances of proposed GF based MFCC and IMFCC in
individual and fused mode have been verified in two standard
databases YOHO, (Microphone Speech) and POLYCOST
(Telephone Speech) each of which has more than 130 speakers.
Abstract: A state of the art Speaker Identification (SI) system requires a robust feature extraction unit followed by a speaker modeling scheme for generalized representation of these features. Over the years, Mel-Frequency Cepstral Coefficients (MFCC) modeled on the human auditory system has been used as a standard acoustic feature set for SI applications. However, due to the structure of its filter bank, it captures vocal tract characteristics more effectively in the lower frequency regions. This paper proposes a new set of features using a complementary filter bank structure which improves distinguishability of speaker specific cues present in the higher frequency zone. Unlike high level features that are difficult to extract, the proposed feature set involves little computational burden during the extraction process. When combined with MFCC via a parallel implementation of speaker models, the proposed feature set outperforms baseline MFCC significantly. This proposition is validated by experiments conducted on two different kinds of public databases namely YOHO (microphone speech) and POLYCOST (telephone speech) with Gaussian Mixture Models (GMM) as a Classifier for various model orders.