Search

In this paper, a sequence-to-sequence model is proposed for translating sentences without idiom to sentence with the idiom. The problem is challenging in two ways, predicting the correct idiom based on context and generating the correct sentence using the idiom due to complex semantic and syntactic rules of language. Sequence-to-sequence learning has gained popularity in the past few years due to their surprising results on machine translation task. This work is based on sequence-to-sequence learning of word sequences along with their part-of-speech tags to predict sentences with correct idiomatic phrases. Results have shown that models have achieved higher BLEU score on using part-of-speech tags as the input sequences. These observations show the prominence of part-of-speech tags in identifying the hidden writing patterns in the language.

Trajectory and performance of launch vehicles are significantly affected by various flight parameters related to propulsion, aerodynamic forces, mass, center of gravity, etc. Precise estimation of these parameters during post-flight analysis is essential for continuous performance improvement of launch vehicles. Since a large number of parameters affect the performance of a launch vehicle, the estimation process becomes computationally complex and time consuming. Estimation of these parameters can be automated by using 6D simulation software which works in tandem with a heuristic method. In this study, an integrated computational procedure has been designed, which employs a 6D simulation software and a genetic algorithm-based optimizer, to estimate the input flight parameters from the actual flight data. The parameters to be estimated, population size, parameter bounds and genetic algorithm parameters such as crossover probability and mutation probability are taken as inputs to the estimation program along with flight trajectory profiles. The objective function to be minimized is derived by computing the squared error between the actual flight trajectory profiles and the simulated trajectory profiles. Since flight measured data contains noise, data smoothing techniques are implemented before usage to avoid trap down in local minima during optimization. Further, to reduce computational time, a relation matrix between input parameters and output flight profiles is devised to intelligently prioritize the parameter estimation. This technique is demonstrated against known input parameter variations and the estimated outputs are found to be very effective.

This paper aims at enhancing the thermo-fluid performance of a plain fin-and-tube heat exchanger. Influence of the height of rectangular type vortex generator on the heat exchanger performance has been analysed. In dimensionless form, the heat transfer is expressed as Colburn factor (j) and pressure drop in terms of the friction factor (f). The arrangement of all the vortex generators follows a “common-flow-up” configuration which tends to counter-rotate the vortices in the wake region of the tubes, thereby, enhancing the thermal mixing to improve the heat transfer. Another parameter, called quality factor (a ratio of j and f), is also evaluated to compare the results between different heights of the winglets. It has been observed that the j factor is the highest for the height 2.8 mm and f factor is the lowest for the height 0.7 mm, whereas the quality factor suggests the height of 0.7 and 1.4 mm to be the appropriate ones.

In this paper, the architecture for idiom recommendation and its implementation based on the syntactical structure of phrases in sentences is proposed. Idioms are a bunch of words that have figurative meaning if taken together, which is different from the actual meaning of individual elements. The entire methodology is divided into four segments—segmentation, analysis, extraction and transformation. Emphasis on every possible syntactic structure of phrases is laid, that indicates similar implication to the structures in the dictionary and recommends idioms based on that.

Face Recognition is immensely proliferating as a research area in the paradigm of Computer Vision as it provides an extensive choice of applications in surveillance and commercial domains. This paper throws light upon the comparison of various dense feature descriptors (Dense SURF, Dense SIFT, Dense ORB) with each other and also with their classical counterparts (SURF, SIFT, ORB) using a novel technique for recognition. This proposed technique uses Laplacian of Gaussian filter for enhancement of the image. It applies various dense and classical feature descriptors on the enhanced image and outputs a feature vector. In order to achieve high performance, this feature vector is given to Fisher vector since Fisher Vector is a feature patch-aggregation method. Finally, extended nearest neighbor Classifier is used for classification over the orthodox k-nearest classifier. Experiments were carried out on three diverse datasets—ORL, Faces94, and Grimace. On scrutinizing the results, Dense SIFT and Dense ORB were found to be preeminent as measured by various performance metrics. 98.44 on Grimace, 98.15 on Faces94.