Papers - THI THI ZIN
-
An Innovative Deep Machine for Human Behavior Analysis Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Proc. of 12th Intl. Conf. on Innovative Computing, Information and Control (ICICIC2017) 2017.8
Language：English Publishing type：Research paper (international conference proceedings)
In this paper we shall propose an innovative deep machine model to analyze human behaviors without using engineers’ hand crafted features. Due to the tremendous potentials with respect to variety of applications and theoretical point of views the deep machine learning techniques and human behavior analysis have become very important and promising two research areas in these days. We shall be bridging these two research areas to gain more benefits in application wise. In particular, we shall introduce two methods: multilevel neural networks with a series of Gamma Activation Functions and deep convolution neural networks using Markov correlated convolution kernels and variable pooling techniques. In addition, back propagation will be used for parameter optimization in order to minimize the mean square errors. The validity of proposed approach will be confirmed by using available public dataset and giving experimental results for comparisons with the best existing results.
-
Human identification using X-Ray image matching Reviewed
Ryudo ISHIGAMI, Thi Thi Zin, Norihiro SHINKAWA, Ryuichi NISHII
Lecture Notes in Engineering and Computer Science 2227 415 - 418 2017.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Engineering and Computer Science
Human identification both prior to death and after death is becoming one of the major worldwide issues nowadays for law enforcement aspect as well as social and security aspects. General identification prior to death is possible through comparison of many biometric identifiers. However identification after death is impossible using behavioral biometric such as speech and actions. Moreover, in many circumstances such as natural disasters, air plane crash or a case of identification a couple of weeks later, most of physical biometrics may not be useful for identification due to the decay of some soft tissues. A lot of research has been done in the field of different biometric modalities like Finger-print, Iris, Hand-Veins, Dental biometrics etc. to identify humans. However only a little has been known the chest X-Ray biometric which was very powerful method for identification especially during the mass disasters in which most of other biometrics are unidentifiable. Therefore in this paper, we propose a stochastic modelling approach for human identification after death by using chest X- Ray prior to death database. Some experimental results are shown based on real life dataset and confirmed.
-
Visual Monitoring System for Elderly People Daily Living Activity Analysis Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Lecture Notes in Engineering and Computer Science 2227 140 - 142 2017.3
Language：English Publishing type：Research paper (international conference proceedings) Publisher：Lecture Notes in Engineering and Computer Science
In today modern world, due to the increasing number of older and vulnerable people better and new approaches are needed to support people in their own homes. Especially, intelligent visual monitoring system recognizes the activities of elderly people daily living become more and more important. This paper proposes a new approach for detecting the daily activities of elderly people whether they are deviated from normal behavior. In this context we will focus on building stochastic models of behavior based on types of activity. Models are trained using only normal behavior. Variations from the models are considered as abnormal behaviors and these can be highlighted for subsequent review or intervention. Experimental results are shown by using some real life datasets to illustrate the proposed models.
-
An effective method for detecting snatch thieves in video surveillance Reviewed
Hiroaki Tsushita, Thi Thi Zin
Proceeding of The 2017 International Conference on Artificial Life and Robotics (ICAROB 2017) 303 - 306 2017.1
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://alife-robotics.co.jp/members2017/icarob/data/html/data/OS_pdf/OS20/OS20-1.pdf
-
Color and Shape based Method for Detecting and Classifying Card Images Reviewed
Cho Nilar Phyo, Thi Thi Zin, Hiroshi Kamada, Takashi Toriu
Proceeding of The 2017 International Conference on Artificial Life and Robotics (ICAROB 2017) 307 - 310 2017.1
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://alife-robotics.co.jp/members2017/icarob/data/html/data/OS_pdf/OS20/OS20-2.pdf
-
Automatic Assessing Body Condition Score from Digital Images by Active Shape Model and Multiple Regression Technique Reviewed
Nay Chi Lynn, Thi Thi Zin, Ikuo Kobayashi
Proceeding of The 2017 International Conference on Artificial Life and Robotics (ICAROB 2017) 311 - 314 2017.1
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://alife-robotics.co.jp/members2017/icarob/data/html/data/OS_pdf/OS20/OS20-3.pdf
-
Block Based Approach for Key Frame Extraction on Large Video Sequences Reviewed
Mie Mie Tin, Nyein Nyein Myo, Mie Mie Khin, Thi Thi Zin
ICIC express letters. Part B, Applications : an international journal of research and surveys 7 ( 12 ) 2713 - 2717 2016.12
Language：English Publishing type：Research paper (scientific journal)
Other Link： http://www.ijicic.org/elb-7(12).htm
-
A study of cow detection and extraction using feature of contrast rate Reviewed
Kosuke Sumi, Thi Thi Zin, Thu Zar Tint, Tin Myint Naing
International Journal of Research in Advanced Engineering and Technology 2 ( 6 ) 5 - 7 2016.11
Language：English Publishing type：Research paper (scientific journal)
In this day, people who become livestock farmers that workload is big are decreasing. When the cow especially acts the calving behavior, livestock farmers need to watch it for a long time. Therefore, the research is aimed for developing cow calving monitoring system to reduce burden of livestock. In the cow calving monitoring system, cow detection and extraction play a vital role. In this paper, the cow detection and extraction are focused on. In order to implement it, image obtained from video camera was analyzed and identified by using techniques of image processing. In the paper, black-haired cows from video sequences are detected and extracted based on inter-frame difference method and contrast rate features. Sobel edge detector and morphological operation are employed to complete results. The experimental results show strong points and weak points of the system.
Other Link： http://www.newengineeringjournal.in/archives/2016/vol2/issue6
-
Deep Learning Model for Integration of Clustering with Ranking in Social Networks Reviewed
Thi Thi Zin, Pyke Tin, Hiromitsu Hama
Genetic and Evolutionary Computing. ICGEC 2016. Advances in Intelligent Systems and Computing, vol 536. Springer 536 247 - 254 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
-
Automatic Target Tracking System based on Local and Global Features Reviewed
Thi Thi Zin, Kenshiro Yamada
Genetic and Evolutionary Computing. ICGEC 2016. Advances in Intelligent Systems and Computing, vol 536. Springer 536 255 - 262 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
-
Reliability and Availability Measures for Internet of Things Consumer World Perspectives Reviewed
Thi Thi Zin, H. Hama, Pyke Tin
Proc. of The 5th 2016 IEEE Global Conf. on Consumer Electronics (GCCE2016) 367 - 368 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1109/GCCE.2016.7800446
Other Link： http://ieeexplore.ieee.org/document/7800446/
-
User-Intent Visual Information Ranking System Reviewed
Swe Nwe Nwe Htun, Thi Thi Zin, M. Yokota, Khin Mo Mo Tun
Proc. of The 5th 2016 IEEE Global Conf. on Consumer Electronics (GCCE2016) 15 - 16 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1109/GCCE.2016.7800316
Other Link： http://ieeexplore.ieee.org/document/7800316/
-
Shape descriptor for binary image retrieval Reviewed
Moe Zet Pwint; Thi Thi Zin; Mitsuhiro Yokota; Mie Mie Tin
Proc. of The 5th 2016 IEEE Global Conf. on Consumer Electronics (GCCE2016) 365 - 366 2016.10
Language：English Publishing type：Research paper (international conference proceedings)
DOI： 10.1109/GCCE.2016.7800445
Other Link： http://ieeexplore.ieee.org/document/7800445/
-
Markov Chain based Query Classification for User Intent Image Search Engines Reviewed
Thi Thi Zin, Pyke Tin and H. Hama
ICIC Express Letters 10 ( 9 ) 2129 - 2134 2016.9
Language：English Publishing type：Research paper (scientific journal)
-
Markov chain based query classification for user intent image search engines
Zin T., Tin P., Hama H.
ICIC Express Letters 10 ( 9 ) 2129 - 2134 2016.9
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters
© 2016 ICIC International. Today Web image search engines have well recognized that a challenging and difficult problem of determining the user intent of Web searches should be thoroughly investigated. In this aspect we propose a Markov chain query classification method to understand user behavior and intent based on the queries which users used for searches. Specifically, we analyze some samples of queries from different Web search engines and classify the user queries into three classes such as informational, navigational and transactional classes. By assuming these three classes represent the type of contents that a user desired to express his or her intents, we implement the Markov chain based classification process. Some experimental results are shown to capture user intents by using a set of queries submitted to the Web search engines.
-
The Identification of Dairy Cows Using Image Processing Techniques Reviewed
Thi Thi Zin, S. Sakurai, K. Sumi, I. Kobayashi, H. Hama
ICIC Express Letters, Part B: Application 7 ( 8 ) 1857 - 1862 2016.8
Language：English Publishing type：Research paper (scientific journal)
-
The identification of dairy cows using image processing techniques
Zin T., Sakurai S., Sumi K., Kobayashi I., Hama H.
ICIC Express Letters, Part B: Applications 7 ( 8 ) 1857 - 1862 2016.8
Language：Japanese Publishing type：Research paper (scientific journal) Publisher：ICIC Express Letters, Part B: Applications
© 2016 ICIC International. Systems to recognize and identify individual cows have become more and more important for modern dairy farms to make precise and successful dairy management. Due to this, various types of identification systems have been coming to front line literature. Most of cow identification systems utilize electrical identification methods based on the RFID (radio frequency identification). However, RFID implementations can cause several problems. Therefore, in this paper we propose an effective and easy-to- use cow identification system by using image processing techniques. Specifically, the proposed system involves recognition of individual cow by identifying the landmark patterns on the cow body along with background modeling, and foreground marks extraction tuned on morphological operations. In order to confirm the system developed in this paper, we present some experimental results based on self-collected video sequences taken at Sumiyoshi Field Science Center, the University of Miyazaki.
-
Action Recognition System with the Microsoft KinectV2 using a Hidden Markov Model Reviewed
M. Fujino, Thi Thi Zin
Proc. of The Third Intl. Conf. on Computing Measurement Control and Sensor Network (CMCSN-2016) 118 - 121 2016.5
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://ieeexplore.ieee.org/document/8008654/
-
A General Video Surveillance Framework for Cow Behavior Analysis Reviewed
Thi Thi Zin, I. Kobayashi, Pyke Tin, H. Hama
Proc. of The Third Intl. Conf. on Computing Measurement Control and Sensor Network (CMCSN-2016) 130 - 133 2016.5
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://ieeexplore.ieee.org/document/8008657/
-
Estrus Detection for Dairy Cow Using a Laser Range Sensor Reviewed
Thi Thi Zin, H. Kai, K. Sumi, I. Kobayashi, H. Hama
Proc. of The Third Intl. Conf. on Computing Measurement Control and Sensor Network (CMCSN-2016) 162 - 165 2016.5
Language：English Publishing type：Research paper (international conference proceedings)
Other Link： http://ieeexplore.ieee.org/document/8008665/