[8] Article “Persian phonemes recognition using PPNet” (Submitted to the journal of "Signal, Image and Video Processing (SIVP)" - [EMID:0ea11c387bec384b])

Abstract:

In this paper a new approach for recognition of Persian phonemes on the PCVC speech dataset is proposed. Nowadays deep neural networks are playing main rule in classification tasks. However the best results in speech recognition are not as good as human recognition rate yet. Deep learning techniques are shown their outstanding performance over so many classification tasks like image classification, document classification, etc. Also in some tasks their performance were even better than human. So the reason why ASR (automatic speech recognition) systems are not as good as the human speech recognition system is mostly depend on features of data is fed to deep neural networks.

In this research first sound samples are cut for exact extraction of phoneme sounds in 50ms samples. Then phonemes are grouped in 30 groups; Containing 23 consonants, 6 vowels and a silence phoneme. STFT (Short time Fourier transform) is applied on them and Then STFT results are given to PPNet (A new deep convolutional neural network architecture) classifier and a total average of 75.87% accuracy is reached which is the best result ever compared to other algorithms on Separated Persian phonemes (Like in PCVC speech dataset).

_____________________________________

[7] Article “Web-pages Semi-Supervised online clustering using deep GRU neural networks” (Preparing to submit)

Abstract:

Today, one of the most important issues in computer science is the analysis of individual behaviors on the Internet and virtual communities. In these systems, computers are tried to understand the behaviors of others as well as humans. The thesis attempts to use a text processing and classification algorithm with deep learning a number of posts in web pages to analyze and to determine if there is a subjective patterns within the texts, and then examine to what extent these topics are closely aligned or separated. For this purpose, the Reuters data collection has been used in which different text structures and separate words are examined. On the training data, classifications were also made on experimental data so artificial neural networks could distinguish texts with different titles.In many cases, web searches are used.

If there were a good algorithm for web page classification, then web pages can be grouped together in the shortest time possible for users to provide optimal search results. Therefore, in academic environments, hospitals, etc., classification of information in a particular area can be used.

_____________________________________

[6] Article “Recognition of negative or positive and sexuality patterns in text using deep GRU neural networks” (Preparing to submit)

Abstract:

Today, one of the most important issues in computer science is the analysis of individual behaviors on the Internet and virtual communities. In these systems, computers are tried to understand the behaviors of others as well as humans. The thesis attempts to use the text processing and classification algorithms as a deep learning method for a number of posts of Different blogs to identify the gender of their writers and have been analyzed to determine whether the gender of the people is effective on their negative and positive emotions, and then examine what types of material each of the two sexes in the different mental states publish? For this purpose, a dataset of positive and negative words was first collected and then a number of texts were selected from the blog post data as the most negative and positive samples. From these data, classification was made on rest of data so artificial neural network can distinguish between positive and negative texts. Then another classroom operation was conducted to identify the gender of the author. Ultimately, classifying operations on gender data for men and women to determine the negative or positive nature of the texts written by each of these sexes determines how positive or negative the difference is between men and women.

_____________________________________

[5] Article “Persian vowel phoneme recognition with MFCC and ANN on PCVC speech dataset” Accepted in “The 5th International Conference of Electrical Engineering, Computer Science and Information Technology 2018”. (Download)

Abstract:

In this paper a new method for recognition of consonant-vowel phonemes combination on a new Persian speech dataset titled as PCVC (Persian Consonant-Vowel Combination) is proposed which is used to recognize Persian phonemes. In PCVC dataset, there are 20 sets of audio samples from 10 speakers which are combinations of 23 consonant and 6 vowel phonemes of Persian language. In each sample, there is a combination of one vowel and one consonant. First, the consonant phoneme is pronounced and just after it, the vowel phoneme is pronounced. Each sound sample is a frame of 2 seconds of audio. In every 2 seconds, there is an average of 0.5 second speech and the rest is silence. In this paper, the proposed method is the implementations of the MFCC (Mel Frequency Cepstrum Coefficients) on every partitioned sound sample. Then, every train sample of MFCC vector is given to a multilayer perceptron feed-forward ANN (Artificial Neural Network) for training process. At the end, the test samples are examined on ANN model for phoneme recognition. After training and testing process, the results are presented in recognition of vowels. Then, the average percent of recognition for vowel phonemes are computed.

_____________________________________

[4] Article “Laser Deviation measurement in passing through objects with pixel coordination changes measurement” (Submitted) (in collaboration with Laser lab of Vali-e-asr university)

_____________________________________

[3] Article “Fuzzy Controller of Reward of Reinforcement Learning in Digit Recognition” accepted in “The third international conference of applied researches in Computer Science and Information Technology 2016” and indexed in Civilica and ISC (In Persian) (Certificate(Civilica Webpage) (Download)

Abstract:

Recognition of human environment with computer systems always was a big deal in artificial intelligence. In this area handwriting recognition and conceptualization of it to computer is an important area in it. In the past years with growth of machine learning in artificial intelligence, efforts to using this technique increased. In this paper is tried to using fuzzy controller, to optimizing amount of reward of reinforcement learning for recognition of handwritten digits. For this aim first a sample of every digit with 10 standard computer fonts, given to actor and then actor is trained. In the next level is tried to test the actor with dataset and then results show improvement of recognition when using fuzzy controller of reinforcement learning.

 _____________________________________

[2] Article “A fast and secure transition method by combination of AES and LZ4 algorithms” accepted in “The third international conference of applied researches in Computer Science and Information Technology 2016” and indexed in Civilica and ISC (In Persian) (Certificate(Civilica Webpage). (Download)

Abstract:

From a long time ago, beside encryption of data and making it secure, compression packing it was also important that could make transmission of data faster. In the past years need for improvement of encryption and compression for a fast and easy transmission is more necessary. In this paper, a new method for combination of LZ4 combination and AES encryption algorithms for a fast and easy packing, securing and compressing of data is presented. Choose of these two algorithms was for some special features of them about aim of this paper. This paper also is introducing a method for Parallelism of compression and encryption in a special way for improvement of speed and security of data.

 _____________________________________

[1] B.S thesis entitled “Design and implementing of process self-protection tool” 2014. (In Persian) (Download)