FCS PuBlicationS

Su Su Aung  | Zin Mar Kyu

Abstract

Background subtraction method is widely used in most of the video motion detection algorithms especially for video surveillance application. Background subtraction is used to extract moving or static foreground objects from the background scene. The efficiency of foreground-background segmentation heavily relies on background model which must be able to cope with changes in the scene and granularity of the foreground objects. The robust background model can produce good foreground segmentation results and it is still a great challenge to get accurate and high performance result today. In this paper, a video foreground-background segmentation approach is proposed. This approach is based on Codebook (CB) model with Kalman Filter. This approach can be used to extract foreground objects from the video stream. The Lab color space is used in this approach to calculate color difference between two pixels using CIEDE2000 color difference formula. Extracted foreground object from video sequence using this approach is useful for object detection in video surveillance applications. This approach has been tested with PETS and CDnet2014 datasets and segmentation results accuracy are evaluated compare with ground truth.

Date: 28th – 29th July, 2017, | IEEE International Conference on Signal Processing and Communication (ICSPC’17), Karunya, India. pp. 332-336| Weblink

Su Su Aung  | Zin Mar Kyu

Abstract

Foreground object extraction is an important subject for computer vision applications. The separation of foreground objects form the background is the crucial step in application such as video surveillance. In order to extract foreground object from a video scene, a background model which can represent dynamic changes in the scene is required. A robust, accurate and high performance approach is still a great challenge today. In this paper, the background modeling approach based on Codebook model with Kalman Filter is presented. This approach can be used to extract foreground objects from the video stream. The Lab color space is used in this approach to calculate color difference between two pixels using CIEDE2000 color difference formula. extracted foreground object from video sequence using this approach is useful for object detection in video surveillance applications.

Date: 16th – 17th February, 2017 | 15th International Conference on Computer Applications (ICCA 2017), Yangon, Myanmar, pp. 312-317| Weblink

Si Si Mar Win | Hnin Mya Aye

Abstract

Mobile Cloud Computing (MCC) provides a platform where mobile users make use of cloud services on mobile devices. The use of mobile clouds in educational settings can provide great opportunities for students as well as researchers to improve their learning outcomes and minimizes the performance, compatibility, and lack of resources issues in mobile computing environment. This paper proposed a MCC based learning model to create an effective learning environment for both the University and learners by integrating virtualized private cloud technology with two components, social networking and mobile learning applications. This model will construct by using High Performance Computing Cloud Toolkit Ezilla and ubiquitous mobile learning elements. It will evaluate by UTAUT based model and Quality of Experience (QoE).

Date: 5th February, 2015 |13th International Conference on Computer Applications (ICCA 2015) | Weblink

Si Si Mar Win  | Hnin Mya Aye

Abstract

Real time communication applications including Mobile learning application can be integrated with other software applications into one platform and deployed in private clouds to reduce capital expenditure and lower overall costs of daily based maintenance and real estate required for computer  hardware. As a critical component of private clouds, virtualization may adversely affect a real time communication application running in virtual machines as the layer of virtualization on the physical server adds system overhead and contributes to capacity lose. Virtualization in the mobile can enable hardware to run with less memory and fewer chips, reducing costs and increasing energy efficiency as well. It also helps to address safety and security challenges, and reduces software development and porting costs. This study will investigate how to build an effective learning environment for both the University and learners by integrating the virtualization, private cloud technology and mobile learning applications.

Date: 26 th August, 2015 | 9thInternational Conference on Genetic and Evolutionary Computing| Weblink

Si Si Mar Win  | Than Nwe Aung

Abstract

Nowadays, text annotation plays an important role within real-time social media mining. Social media analysis provides actionable information to its users in times of natural disasters. This paper presents an approach to a real-time two layer text annotation system for social media stream to the domain of natural disasters. The proposed system annotates raw tweets from Twitter into two types such as Informative or Not Informative as first layer. And then it annotates again five information types based on Informative tweets only as second layer. Based on the first and second layer annotation results, this system provides the tweets with user desired informative type in real time. In this system, annotation is done at tweet level by using word and phrase level features with LibLinear classifier. All features are in the form of Ngram nature based on part of speech (POS) tag, Sentiment Lexicon and especially created Disaster Lexicon. The validation of this system is performed based on different disaster related datasets and new Myanmar_Earthquake_2016 dataset derived from Twitter. The annotated datasets generated from this work can also be used by interested research communities to study the social media natural disaster related research.

Date: 12th March, 2018 | Advances in Science, Technology and Engineering Systems Journal Vol. 3, No. 2, 119-127 (2018)| Weblink

Si Si Mar Win  | Than Nwe Aung

Abstract

Nowadays, text annotation plays an important role within real-time social media mining. Social media analysis provides actionable information to its users in times of natural disasters. This paper presents an approach to a real-time two layer text annotation system for social media stream to the domain of natural disasters. The proposed system annotates raw tweets from Twitter into two types such as Informative or Not Informative as first layer. And then it annotates again five information types based on Informative tweets only as second layer. Based on the first and second layer annotation results, this system provides the tweets with user desired informative type in real time. In this system, annotation is done at tweet level by using word and phrase level features with LibLinear classifier. All features are in the form of Ngram nature based on part of speech (POS) tag, Sentiment Lexicon and especially created Disaster Lexicon. The validation of this system is performed based on different disaster related datasets and new Myanmar_Earthquake_2016 dataset derived from Twitter. The annotated datasets generated from this work can also be used by interested research communities to study the social media natural disaster related research.

Date: 16th  February, 2017 | 15th International Conference On Computer Applications (ICCA 2017)| Weblink

Si Si Mar Win  | Than Nwe Aung

Abstract

Nowadays, text annotation plays an important role within real-time social media mining. Social media analysis provides actionable information to its users in times of natural disasters. This paper presents an approach to a real-time two layer text annotation system for social media stream to the domain of natural disasters. The proposed system annotates raw tweets from Twitter into two types such as Informative or Not Informative as first layer. And then it annotates again five information types based on Informative tweets only as second layer. Based on the first and second layer annotation results, this system provides the tweets with user desired informative type in real time. In this system, annotation is done at tweet level by using word and phrase level features with LibLinear classifier. All features are in the form of Ngram nature based on part of speech (POS) tag, Sentiment Lexicon and especially created Disaster Lexicon. The validation of this system is performed based on different disaster related datasets and new Myanmar_Earthquake_2016 dataset derived from Twitter. The annotated datasets generated from this work can also be used by interested research communities to study the social media natural disaster related research.

Date:  1st July, 2017 | International Journal of Networked and Distributed Computing, Vol. 5, No. 3 (July 2017) 133–142

Si Si Mar Win  | Than Nwe Aung

Abstract

Nowadays, text annotation plays an important role within real-time social media mining. Social media analysis provides actionable information to its users in times of natural disasters. This paper presents an approach to a real-time two layer text annotation system for social media stream to the domain of natural disasters. The proposed system annotates raw tweets from Twitter into two types such as Informative or Not Informative as first layer. And then it annotates again five information types based on Informative tweets only as second layer. Based on the first and second layer annotation results, this system provides the tweets with user desired informative type in real time. In this system, annotation is done at tweet level by using word and phrase level features with LibLinear classifier. All features are in the form of Ngram nature based on part of speech (POS) tag, Sentiment Lexicon and especially created Disaster Lexicon. The validation of this system is performed based on different disaster related datasets and new Myanmar_Earthquake_2016 dataset derived from Twitter. The annotated datasets generated from this work can also be used by interested research communities to study the social media natural disaster related research.

Date: 24th May, 2017 | IEEE/ACIS 16th International Conference on Computer and Information Science (ICIS) | Weblink

Pyae Phyo Thu  |  Htwe Nu Win

Abstract

The term virtual phone conversation system refers to the bidirectional Myanmar to Myanmar Speech-to-Text and Text-to-Speech translation system that runs in real-time on the mobile phones. With the used of Real-time Speech Recognizer and Real-time Speech Synthesizer, this bidirectional phone  conversation translation capability helps the dumb and the hearing impaired person to speak on the mobile phone like the normal person. In this paper, we will propose the framework of the system, its work flow, its applicable methodologies and its challenges.

Date: 1st February, 2015 |  13th International Conference on Computer Applications (ICCA 2015) | Weblink

Pyae Phyo Thu  |  Nwe Nwe

Abstract

Due to the implicit traits embedded in tweets, handling figurative languages appear as the most trending topics in computational linguistics. While recognition of a single language is hard to capture, differentiating several languages at once is the most challenging task. To achieve this purpose, we employ a set of emotion-based features in order to individuate between humor, irony, sarcasm, satire and true. We use eight basic emotions excerpted from EmoLex supplement with tweets polarity. We apply these features in two datasets: balanced dataset (collected using hashtag-based approach) and class-imbalanced dataset (collected from streaming tweets). As a result, the model not only outperform a word-based baseline but also handle both balanced and class-imbalanced datasets in multi-figurative language detection.

Date: 1st  May, 2017 | In Computer and Information Science (ICIS), 2017 IEEE/ACIS 16th International Conference on (pp. 209-214). IEEE  | Weblink

Pyae Phyo Thu  |   Than Nwe Aung

Abstract

Recognition of satirical language in social multimedia outlets turns out to be a trending research area in computational linguistics. Many researchers have analyzed satirical language from the various point of views: lexically, syntactically, and semantically. However, due to the ironic dimension of emotion embedded in the language, emotional study of satirical language has ever left behind. This paper proposes the emotion-based detection system for satirical figurative language processing. These emotional features are extracted using emotion lexicon: EmoLex and sentiment lexicon: VADER. Ensemble bagging technique is used to tackle the problem of ambiguous nature of emotion. Experiments are carried out on both short text and long text configurations namely news articles, Amazon product reviews, and tweets. Recognition of satirical language can aid in lessening the impact of implicit language in public opinion mining, sentiment analysis, fake news detection and cyberbullying.

Date:   26-28th  June, 2017 | Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD), 2017 18th IEEE/ACIS International Conference on (pp. 149-154). IEEE  | Weblink

Pyae Phyo Thu  |   Than Nwe Aung

Abstract

Even though the various features of satirical language have been studied in computational linguistics, most of the research works have relied on the performance of the single machine learning algorithm. However, the implicit traits embedded in the language demand more certain, precise and accurate combination powers of an individual algorithm. In this study, we analyzed the performance of emotion-based satire detection model on various machine learning algorithms: Regression, Naïve Bayes, SVM and ensemble classifiers. Experiments on shifting base classifiers to ensemble classifiers demonstrate that ambiguous and implicit nature of satirical emotions can lead to the misclassification accuracy while implementing the base classifiers but, offer reliable classification accuracy with ensemble classifiers.

Date:   24th    October, 2017 | Consumer Electronics (GCCE), 2017 IEEE 6th Global Conference on (pp. 1-5). IEEE  | Weblink

Pyae Phyo Thu  |   Than Nwe Aung

Abstract

Recognition of satirical language in social multimedia outlets turns out to be a trending research area in computational linguistics. Many researchers have analyzed satirical language from the various point of views: lexically, syntactically, and semantically. However, due to the ironic dimension of emotion embedded in the language, emotional study of satirical language has ever left behind. This paper proposes the emotion-based detection system for satirical figurative language processing. These emotional features are extracted using emotion lexicon: EmoLex and sentiment lexicon: VADER. Ensemble bagging technique is used to tackle the problem of ambiguous nature of emotion. Experiments are carried out on both short text and long text configurations namely news articles, Amazon product reviews, and tweets. Recognition of satirical language can aid in lessening the impact of implicit language in public opinion mining, sentiment analysis, fake news detection and cyberbullying.

Date:   30th April, 2018 | International Journal of Networked and Distributed Computing (IJNDC), Vol. 6, No. 2 (April 2018), pp.78-87, ISSN: 2211-7946, DOI: doi:10.2991/ijndc.2018.6.2.3 | Weblink

Pyae Phyo Thu  |   Than Nwe Aung

Abstract

Due to the implicit traits embedded in the language, handling figurative languages appear to be the most trending topics in public opinion mining and social multimedia sentiment analysis. Failures in recognition of these languages can lead to the misrepresentation of actual sentiments, attitudes or opinions person or community try to expose. Satire is a more alive form of figurative communication which intends to criticize someone’s behavior and ridicule it. This work proposes the POS level emotion-based features by using the emotion lexicon SenticNet and VADER. It is approached as a classification problem by applying a supervised machine learning algorithm: Random Forest. The system can tackle the problem of high bias error in both long text and short text datasets with 83% to 89% accuracy whereas the BOW gives high accuracy but cannot handle the problem of high bias error in satirical language processing.

Date:   1st December, 2018 | International Journal of Computer Science and Network( IJCSN), Volume 7, Issue 6, December 2018 | Weblink