Gearbox is one of the vital components in aircraft engines. If any small damage to gearbox, it can cause the breakdown of aircraft engine. Thus it is significant to study fault diagnosis in gearbox system. In this paper, two deep learning models (Long short term memory (LSTM) and Bi-directional long short term memory (BLSTM)) are proposed to classify the condition of gearbox into good or bad. These models are applied on aircraft gearbox vibration data in both time and frequency domain. A publicly available aircraft gearbox vibration dataset is used to evaluate the performance of proposed models. The results proved that accuracy achieved by LSTM and BLSTM are highly reliable and applicable in health monitoring of aircraft gearbox system in time domain as compared to frequency domain. Also, to show the superiority of proposed models for aircraft gearbox fault diagnosis, performance is compared with classical machine learning models.

a b s t r a c tIn this work, we investigate the effect of texture features for the classification of flower images. A flower image is segmented by eliminating the background using a thresholdbased method. The texture features, namely the color texture moments, gray-level co-occurrence matrix, and Gabor responses, are extracted, and combinations of these three are considered in the classification of flowers. In this work, a probabilistic neural network is used as a classifier. To corroborate the efficacy of the proposed method, an experiment was conducted on our own data set of 35 classes of flowers, each with 50 samples. The data set has different flower species with similar appearance (small inter-class variations) across different classes and varying appearance (large intra-class variations) within a class. Also, the images of flowers are of different pose, with cluttered background under various lighting conditions and climatic conditions. The experiment was conducted for various sizes of the datasets, to study the effect of classification accuracy, and the results show that the combination of multiple features vastly improves the performance, from 35% for the best single feature to 79% for the combination of all features. A qualitative comparative analysis of the proposed method with other well-known existing state of the art flower classification methods is also given in this paper to highlight the superiority of the proposed method.

a b s t r a c tIn this work, an approach for online signature verification based on writer specific features and classifier is investigated. Existing models for online signatures are generally writer independent, as a common classifier or fusion of classifier is used on a common set of features for all writers during verification. In contrast, our approach is based on the usage writer dependent features as well as writer dependent classifier. The two decisions namely optimal features suitable for a writer and a classifier to be used for authenticating the writer are taken based on the error rate achieved with the training samples. The performance of our model is tested on both MCYT-100 (DB1), a sub corpus of MCYT data set, consisting of signatures of 100 writers, MCYT-330 (DB2) consisting of signatures of all 330 writers and visual subcorpus of SUSIG dataset. Experimental results confirm the effectiveness of writer dependent characteristics for online signature verification. The error rate that we achieved is lower when compared to many existing contemporary works on online signature verification especially when the number of training samples available for each writer is sufficient enough.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Highlights  An approach for online signature verification based on writer dependent parameters  Interval valued symbolic representation of writer dependent features  Verification based on both symbolic representation and conventional representation  Lowest EER with symbolic representation and writer dependent parameters  Obtained results indicate the superiority of the proposed approach

No abstract