S C Premaratne, D D Karunaratna, G N Wikramanayake, K P Hewagamage, G K A Dias (2004) "Profile Based Video Segmentation System to Support e-Learning" In:6th International Information Technology Conference, Edited by:V.K. Samaranayake et al. pp. 74-81. Infotel Lanka Society, Colombo, Sri Lanka: IITC Nov 29-Dec 1, ISBN: 955-8974-01-3
This paper presents a semantic model which delivers personalized audio information. The personalization process is automated and decentralized. The metadata which support personalization are separated in two categories: the metadata describing user preferences stored at each user and the resource adaptation metadata stored at the server. The multimedia models MPEG-21 and MPEG-7 are used to describe metadata information. The Web Ontology Language (OWL) language is used to produce and manipulate the relative ontological descriptions.
This paper presents an audio personalization framework for mobile devices. The multimedia
models MPEG-21 and MPEG-7 are used to describe metadata information. The metadata which support personalization are stored into each device. The Web Ontology Language (OWL) language is used to produce and manipulate the relative ontological descriptions. The process is distributed according to the MapReduce framework and implemented over the Android platform. It determines a hierarchical system structure consisted of Master and Worker devices. The Master retrieves a list of audio tracks matching specific criteria using SPARQL queries.
Content Based Video Retrieval Using Integrated Feature Extraction and Persona...IJERD Editor
Traditional video retrieval methods fail to meet technical challenges due to large and rapid growth of
multimedia data, demanding effective retrieval systems. In the last decade Content Based Video Retrieval
(CBVR) has become more and more popular. The amount of lecture video data on the Worldwide Web (WWW)
is growing rapidly. Therefore, a more efficient method for video retrieval in WWW or within large lecture video
archives is urgently needed. This paper presents an implementation of automated video indexing and video
search in large videodatabase. First of all, we apply automatic video segmentation and key-frame detection to
extract the frames from video. At next, we extract textual keywords by applying on video i.e. Optical Character
Recognition (OCR) technology on key-frames and Automatic Speech Recognition (ASR) on audio tracks of that
video. At next, we also extractingcolour, texture and edge detector features from different method. At last, we
integrate all the keywords and features which has extracted from above techniques for searching
purpose.Finallysearch similarity measure is applied to retrieve the best matchingcorresponding videos are
presented as output from database. Additionally we are providing Re-ranking of results as per users interest in
original result.
This paper presents a semantic model which delivers personalized audio information. The personalization process is automated and decentralized. The metadata which support personalization are separated in two categories: the metadata describing user preferences stored at each user and the resource adaptation metadata stored at the server. The multimedia models MPEG-21 and MPEG-7 are used to describe metadata information. The Web Ontology Language (OWL) language is used to produce and manipulate the relative ontological descriptions.
This paper presents an audio personalization framework for mobile devices. The multimedia
models MPEG-21 and MPEG-7 are used to describe metadata information. The metadata which support personalization are stored into each device. The Web Ontology Language (OWL) language is used to produce and manipulate the relative ontological descriptions. The process is distributed according to the MapReduce framework and implemented over the Android platform. It determines a hierarchical system structure consisted of Master and Worker devices. The Master retrieves a list of audio tracks matching specific criteria using SPARQL queries.
Content Based Video Retrieval Using Integrated Feature Extraction and Persona...IJERD Editor
Traditional video retrieval methods fail to meet technical challenges due to large and rapid growth of
multimedia data, demanding effective retrieval systems. In the last decade Content Based Video Retrieval
(CBVR) has become more and more popular. The amount of lecture video data on the Worldwide Web (WWW)
is growing rapidly. Therefore, a more efficient method for video retrieval in WWW or within large lecture video
archives is urgently needed. This paper presents an implementation of automated video indexing and video
search in large videodatabase. First of all, we apply automatic video segmentation and key-frame detection to
extract the frames from video. At next, we extract textual keywords by applying on video i.e. Optical Character
Recognition (OCR) technology on key-frames and Automatic Speech Recognition (ASR) on audio tracks of that
video. At next, we also extractingcolour, texture and edge detector features from different method. At last, we
integrate all the keywords and features which has extracted from above techniques for searching
purpose.Finallysearch similarity measure is applied to retrieve the best matchingcorresponding videos are
presented as output from database. Additionally we are providing Re-ranking of results as per users interest in
original result.
NL based Object Oriented modeling - EJSR 35(1)IT Industry
Imran Sarwar Bajwa, Shahzad Mumtaz, Ali Samad [2009], "Object Oriented Software Modeling using NLP Based Knowledge Extraction", European Journal of Scientific Research, Aug 2009, Vol. 35 No. 01, pp:22-33
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Content based video retrieval using discrete cosine transformnooriasukmaningtyas
A content based video retrieval (CBVR) framework is built in this paper.
One of the essential features of video retrieval process and CBVR is a color
value. The discrete cosine transform (DCT) is used to extract a query video
features to compare with the video features stored in our database. Average
result of 0.6475 was obtained by using the DCT after implementing it to the
database we created and collected, and on all categories. This technique was
applied on our database of video, check 100 database videos, 5 videos in
Keywords: each category.
Design and Implementation of Efficient Search Methodology for Content-Based R...IDES Editor
E-Learning portal is the full of content of different
formats like text, metadata, image, audio, and video. Current
search methodologies have a direct impact on the fundamental
retrieval issues that information seekers encounter in their use
of the vast number of search systems on the web today. Recently,
information retrieval for text and multimedia content has
become an important research area. Content-based retrieval
in multimedia is a challenging problem since multimedia data
needs detailed interpretation from pixel values. Based on
several new technologies, such as ubiquitous computing,
ontology engineering, semantic web and grid computing, it is
observed that for flexible educational platform architecture
for E-Learning that is OntoEdu is must. In this paper we offer
review report of E-Learning architecture and propose efficient
search algorithm to retrieve multimedia content from the ELearning
environment. The purpose of this technique is to
efficient and fast retrieval of data from content based
environment. The results of these proposed searching
techniques have been found satisfactorily.
A Personalized Audio Web Service using MPEG-7 and MPEG-21 standardsUniversity of Piraeus
This paper presents a web service which
delivers personalized audio information. The personalization process is automated and decentralized. The metadata which support personalization are separated in two categories: the metadata describing user preferences stored at each user and the resource adaptation metadata stored at the web service host. The multimedia models MPEG-21 and MPEG-7 are used to describe metadata information and the Web Ontology Language (OWL) to produce and manipulate ontological descriptions. SPARQL is used for querying the OWL ontologies. The MPEG Query Format (MPQF) is also used, providing a wellknown framework for applying queries to the metadata and to the ontologies.
System analysis and design for multimedia retrieval systemsijma
Due to the extensive use of information technology and the recent developments in multimedia systems, the
amount of multimedia data available to users has increased exponentially. Video is an example of
multimedia data as it contains several kinds of data such as text, image, meta-data, visual and audio.
Content based video retrieval is an approach for facilitating the searching and browsing of large
multimedia collections over WWW. In order to create an effective video retrieval system, visual perception
must be taken into account. We conjectured that a technique which employs multiple features for indexing
and retrieval would be more effective in the discrimination and search tasks of videos. In order to validate
this, content based indexing and retrieval systems were implemented using color histogram, Texture feature
(GLCM), edge density and motion..
NL based Object Oriented modeling - EJSR 35(1)IT Industry
Imran Sarwar Bajwa, Shahzad Mumtaz, Ali Samad [2009], "Object Oriented Software Modeling using NLP Based Knowledge Extraction", European Journal of Scientific Research, Aug 2009, Vol. 35 No. 01, pp:22-33
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Content based video retrieval using discrete cosine transformnooriasukmaningtyas
A content based video retrieval (CBVR) framework is built in this paper.
One of the essential features of video retrieval process and CBVR is a color
value. The discrete cosine transform (DCT) is used to extract a query video
features to compare with the video features stored in our database. Average
result of 0.6475 was obtained by using the DCT after implementing it to the
database we created and collected, and on all categories. This technique was
applied on our database of video, check 100 database videos, 5 videos in
Keywords: each category.
Design and Implementation of Efficient Search Methodology for Content-Based R...IDES Editor
E-Learning portal is the full of content of different
formats like text, metadata, image, audio, and video. Current
search methodologies have a direct impact on the fundamental
retrieval issues that information seekers encounter in their use
of the vast number of search systems on the web today. Recently,
information retrieval for text and multimedia content has
become an important research area. Content-based retrieval
in multimedia is a challenging problem since multimedia data
needs detailed interpretation from pixel values. Based on
several new technologies, such as ubiquitous computing,
ontology engineering, semantic web and grid computing, it is
observed that for flexible educational platform architecture
for E-Learning that is OntoEdu is must. In this paper we offer
review report of E-Learning architecture and propose efficient
search algorithm to retrieve multimedia content from the ELearning
environment. The purpose of this technique is to
efficient and fast retrieval of data from content based
environment. The results of these proposed searching
techniques have been found satisfactorily.
A Personalized Audio Web Service using MPEG-7 and MPEG-21 standardsUniversity of Piraeus
This paper presents a web service which
delivers personalized audio information. The personalization process is automated and decentralized. The metadata which support personalization are separated in two categories: the metadata describing user preferences stored at each user and the resource adaptation metadata stored at the web service host. The multimedia models MPEG-21 and MPEG-7 are used to describe metadata information and the Web Ontology Language (OWL) to produce and manipulate ontological descriptions. SPARQL is used for querying the OWL ontologies. The MPEG Query Format (MPQF) is also used, providing a wellknown framework for applying queries to the metadata and to the ontologies.
System analysis and design for multimedia retrieval systemsijma
Due to the extensive use of information technology and the recent developments in multimedia systems, the
amount of multimedia data available to users has increased exponentially. Video is an example of
multimedia data as it contains several kinds of data such as text, image, meta-data, visual and audio.
Content based video retrieval is an approach for facilitating the searching and browsing of large
multimedia collections over WWW. In order to create an effective video retrieval system, visual perception
must be taken into account. We conjectured that a technique which employs multiple features for indexing
and retrieval would be more effective in the discrimination and search tasks of videos. In order to validate
this, content based indexing and retrieval systems were implemented using color histogram, Texture feature
(GLCM), edge density and motion..
Attendance Management System using Face RecognitionNanditaDutta4
The project ppt presentation is made for the academic session for the completion of the work from Bharati Vidyapeeth Deemed University(IMED) MCA department
Automatic semantic content extraction in videos using a fuzzy ontology and ru...IEEEFINALYEARPROJECTS
To Get any Project for CSE, IT ECE, EEE Contact Me @ 09849539085, 09966235788 or mail us - ieeefinalsemprojects@gmail.co¬m-Visit Our Website: www.finalyearprojects.org
Personalized Multimedia Web Services in Peer to Peer Networks Using MPEG-7 an...University of Piraeus
Multimedia information has been increased in the recent years while new content delivery services enhanced with personalization functionalities are provided to users. Several standards are proposed for the representation and retrieval of multimedia content. This paper makes an overview of the available standards and technologies. Furthermore a prototype semantic P2P architecture is presented which delivers personalized audio information. The metadata which support personalization are separated in two categories: the metadata describing user preferences stored at each user and the resource adaptation metadata stored at the P2P network’s web services. The multimedia models MPEG-21 and MPEG-7 are used to describe metadata information and the Web Ontology Language (OWL) to produce and manipulate ontological descriptions. SPARQL is used for querying the OWL ontologies. The MPEG Query Format (MPQF) is also used, providing a well-known framework for applying queries to the metadata and to the ontologies.
T Silva, D D Karunaratna, G N Wikramanayake, K P Hewagamage, G K A Dias (2004) "Speaker Search and Indexing for Multimedia Databases" In:6th International Information Technology Conference, Edited by:V.K. Samaranayake et al. pp. 157-162. Infotel Lanka Society, Colombo, Sri Lanka: IITC Nov 29-Dec 1, ISBN: 955-8974-01-3
T Silva, D D Karunaratna, G N Wikramanayake, K P Hewagamage, G K A Dias (2004) Speaker Search and Indexing for Multimedia Databases In: 6th International Information Technology Conference Edited by:V.K. Samaranayake et al. 157-162 Infotel Lanka Society Colombo, Sri Lanka: IITC Nov 29-Dec 1, ISBN: 955-8974-01-3
RECURRENT FEATURE GROUPING AND CLASSIFICATION MODEL FOR ACTION MODEL PREDICTI...IJDKP
Content based retrieval has an advantage of higher prediction accuracy as compared to tagging based approach. However, the complexity in its representation and classification approach, results in lower processing accuracy and computation overhead. The correlative nature of the feature data are un-explored in the conventional modeling, where all the data features are taken as a set of feature values to give a decision. The recurrent feature class attribute is observed for the feature regrouping in action model prediction. In this paper a co-relative information, bounding grouping approach is suggested for action model prediction in CBMR application. The co-relative recurrent feature mapping results in faster retrieval process as compared to the conventional retrieval system.
RECURRENT FEATURE GROUPING AND CLASSIFICATION MODEL FOR ACTION MODEL PREDICTI...IJDKP
Content based retrieval has an advantage of higher prediction accuracy as compared to tagging based approach. However, the complexity in its representation and classification approach, results in lower processing accuracy and computation overhead. The correlative nature of the feature data are un-explored in the conventional modeling, where all the data features are taken as a set of feature values to give a decision. The recurrent feature class attribute is observed for the feature regrouping in action model prediction. In this paper a co-relative information, bounding grouping approach is suggested for action model prediction in CBMR application. The co-relative recurrent feature mapping results in faster retrieval process as compared to the conventional retrieval system.
Abstract Recently, Video is becoming a catholic medium for e-learning. As per the popularity of online video information over the World Wide Web (WWW) is mostly dependent on user-assigned tags or specification, which is the system by which we can access such videos. However, this system have limitations for retrieval and frequently we want access to the content (pacify) of the video itself is directly matched against a user’s query except manually assigned tags or specifications. In e-lecturing videos it contains visual and aural medium: slides of presentation and speech. in this system, we are going to retrieve the text from the videos automatically. To abstract visible information, we apply video content analysis to detect slides and optical character recognition to obtain their text. We abstract textual metadata by applying video Optical Character Recognition (OCR) technology on key-frames and Automatic Speech Recognition (ASR) on lecture audio. The ASR and OCR translate and discover slide text line types are accept for keywords abstraction, in which video and fragment-level keywords are abstracted for video searching on the basis of contents. .Key Words: Video fragmentation, Frame Abstraction, video indexing, and etc
Evaluation of English and IT skills of new entrants to Sri Lankan universitiesGihan Wikramanayake
Gihan N. Wikramanayake, Damitha D. Karunartna, Dilkushi S. Wettewe, "Evaluation of English and IT skills of new entrants to Sri Lankan universities", International Conference on Information and Educational Technology (ICIET), Mumbai, 15 Jan 2012.
This study presents our experiences in designing, implementing and deploying an on-line evaluation scheme to measure the English and information technology skills of new entrants to Sri Lankan universities at point of entry in 2011. Over 15,000 students from 25 districts of the country were subjected to the on-line evaluation. The test was
conducted by using a learning management system, in 24 consecutive days in twenty six centres scattered across the country. This paper sums up the experiences we gathered in conducting the evaluation of a larger group of students spread across a wide geographical area and the lessons learned.
G N Wikramanayake (2010) Learning beyond the classroom In: Humanitarian Technology Challenges of the 21st Century, Trivandrum, Kerala, 20-21 Feb. IEEE Kerala Section
Seminar on Sports and Information Technology held at UCSC on 10th July 2010 under the distinguish patronage of Hon. C.B. Rathnayake Minister of Sports, Member of Parliament Thilanga Sumithipala and Professor Kshanika Hirimburegama Vice-Chancellor, University of Colombo
Improving student learning through assessment for learning using social media...Gihan Wikramanayake
Hakim Usoof, Gihan Wikramanayake (2009) Improving student learning through assessment for learning using social media and e-Learning 2.0 on a distance education degree programme in Sri Lanka In: Open Learning: Media, Environments and Cultures, What Role for Social Media and E-Learning 2.0? The European Conference on Educational Research (ECER) in Vienna, Austria: Sept 28-30
M C Siriwardena, G N Wikramanayake (2005) Exploiting Tourism through Data Warehousing IS Engineer, The Bulletin of the British Computer Society Sri Lanka Section, Oct, pp. 23-25.
Authropometry of Sri Lankan Sportsmen and Sportswomen, with Special Reference...Gihan Wikramanayake
T W Wikramanayake, J Dassanayake, G N Wikramanayake, S Amerasinghe (1991) Authropometry of Sri Lankan Sportsmen and Sportswomen, with Special Reference to Body Mass Index The Ceylon Journal of Medical Science 34: 1. 15-32 Jun
Analysis of Multiple Choice Question Papers with Special Reference to those s...Gihan Wikramanayake
V K Samaranayake, G N Wikramanayake, A P S R Somasiri, M G N A S Fernando (1985) Analysis of Multiple Choice Question Papers with Special Reference to those set at the G.C.E. (Advanced Level) Examination The Journal of the Mathematical and Astronomical Society 12: 17-25
P G Punchihewa, G N Wikramanayake, D D Karunaratna (2003) Balanced Scorecard and its relationship to UMM IS Engineer, The Bulletin of the British Computer Society Sri Lanka Section 7-8 Oct
H A Caldera, Y Deshpande, G N Wikramanayake (2005) Web Usage Mining Based on Heuristics: Drawbacks. IS Engineer, The Bulletin of the British Computer Society Sri Lanka Section, Apr, pp. 27-28.
G N Wikramanayake, W A Gray, N J Fiddian (1995) Evolving and Migrating Relational Legacy Databases In:14th Conference of South East Asia Regional Computer Confederation on Sharing IT Achievements for Regional Growth 533-561 Computer Society of Sri Lanka for SEARCC CSSL Sep 5-8, ISBN 955-9155-03-2
Re-Engineering Databases using Meta-Programming TechnologyGihan Wikramanayake
G N Wikramanayake (1997) "Re-engineering Databases using Meta-Programming Technology" In:16th National Information Technology Conference on Information Technology for Better Quality of Life Edited by:R. Ganepola et al. pp. 1-14. Computer Society of Sri Lanka, Colombo: CSSL Jul 11-13, ISBN 955-9155-05-9
P N P Fernando, G N Wikramanayake (1998) "Development of a Web site with Dynamic Data" In: 54th Annual Sessions of Sri Lanka Association for the Advancement of Science, pp. 246-247. Colombo: SLAAS Dec 14-19, Part 1 – abstracts
2024.06.01 Introducing a competency framework for languag learning materials ...Sandy Millin
http://sandymillin.wordpress.com/iateflwebinar2024
Published classroom materials form the basis of syllabuses, drive teacher professional development, and have a potentially huge influence on learners, teachers and education systems. All teachers also create their own materials, whether a few sentences on a blackboard, a highly-structured fully-realised online course, or anything in between. Despite this, the knowledge and skills needed to create effective language learning materials are rarely part of teacher training, and are mostly learnt by trial and error.
Knowledge and skills frameworks, generally called competency frameworks, for ELT teachers, trainers and managers have existed for a few years now. However, until I created one for my MA dissertation, there wasn’t one drawing together what we need to know and do to be able to effectively produce language learning materials.
This webinar will introduce you to my framework, highlighting the key competencies I identified from my research. It will also show how anybody involved in language teaching (any language, not just English!), teacher training, managing schools or developing language learning materials can benefit from using the framework.
June 3, 2024 Anti-Semitism Letter Sent to MIT President Kornbluth and MIT Cor...Levi Shapiro
Letter from the Congress of the United States regarding Anti-Semitism sent June 3rd to MIT President Sally Kornbluth, MIT Corp Chair, Mark Gorenberg
Dear Dr. Kornbluth and Mr. Gorenberg,
The US House of Representatives is deeply concerned by ongoing and pervasive acts of antisemitic
harassment and intimidation at the Massachusetts Institute of Technology (MIT). Failing to act decisively to ensure a safe learning environment for all students would be a grave dereliction of your responsibilities as President of MIT and Chair of the MIT Corporation.
This Congress will not stand idly by and allow an environment hostile to Jewish students to persist. The House believes that your institution is in violation of Title VI of the Civil Rights Act, and the inability or
unwillingness to rectify this violation through action requires accountability.
Postsecondary education is a unique opportunity for students to learn and have their ideas and beliefs challenged. However, universities receiving hundreds of millions of federal funds annually have denied
students that opportunity and have been hijacked to become venues for the promotion of terrorism, antisemitic harassment and intimidation, unlawful encampments, and in some cases, assaults and riots.
The House of Representatives will not countenance the use of federal funds to indoctrinate students into hateful, antisemitic, anti-American supporters of terrorism. Investigations into campus antisemitism by the Committee on Education and the Workforce and the Committee on Ways and Means have been expanded into a Congress-wide probe across all relevant jurisdictions to address this national crisis. The undersigned Committees will conduct oversight into the use of federal funds at MIT and its learning environment under authorities granted to each Committee.
• The Committee on Education and the Workforce has been investigating your institution since December 7, 2023. The Committee has broad jurisdiction over postsecondary education, including its compliance with Title VI of the Civil Rights Act, campus safety concerns over disruptions to the learning environment, and the awarding of federal student aid under the Higher Education Act.
• The Committee on Oversight and Accountability is investigating the sources of funding and other support flowing to groups espousing pro-Hamas propaganda and engaged in antisemitic harassment and intimidation of students. The Committee on Oversight and Accountability is the principal oversight committee of the US House of Representatives and has broad authority to investigate “any matter” at “any time” under House Rule X.
• The Committee on Ways and Means has been investigating several universities since November 15, 2023, when the Committee held a hearing entitled From Ivory Towers to Dark Corners: Investigating the Nexus Between Antisemitism, Tax-Exempt Universities, and Terror Financing. The Committee followed the hearing with letters to those institutions on January 10, 202
Safalta Digital marketing institute in Noida, provide complete applications that encompass a huge range of virtual advertising and marketing additives, which includes search engine optimization, virtual communication advertising, pay-per-click on marketing, content material advertising, internet analytics, and greater. These university courses are designed for students who possess a comprehensive understanding of virtual marketing strategies and attributes.Safalta Digital Marketing Institute in Noida is a first choice for young individuals or students who are looking to start their careers in the field of digital advertising. The institute gives specialized courses designed and certification.
for beginners, providing thorough training in areas such as SEO, digital communication marketing, and PPC training in Noida. After finishing the program, students receive the certifications recognised by top different universitie, setting a strong foundation for a successful career in digital marketing.
Model Attribute Check Company Auto PropertyCeline George
In Odoo, the multi-company feature allows you to manage multiple companies within a single Odoo database instance. Each company can have its own configurations while still sharing common resources such as products, customers, and suppliers.
Macroeconomics- Movie Location
This will be used as part of your Personal Professional Portfolio once graded.
Objective:
Prepare a presentation or a paper using research, basic comparative analysis, data organization and application of economic information. You will make an informed assessment of an economic climate outside of the United States to accomplish an entertainment industry objective.
Unit 8 - Information and Communication Technology (Paper I).pdfThiyagu K
This slides describes the basic concepts of ICT, basics of Email, Emerging Technology and Digital Initiatives in Education. This presentations aligns with the UGC Paper I syllabus.
How to Make a Field invisible in Odoo 17Celine George
It is possible to hide or invisible some fields in odoo. Commonly using “invisible” attribute in the field definition to invisible the fields. This slide will show how to make a field invisible in odoo 17.
Advantages and Disadvantages of CMS from an SEO Perspective
Profile based Video segmentation system to support E-learning
1. Profile based Video segmentation system to support E-learning
S. C. Premaratne, D. D. Karunaratna, G. N. Wikramanayake
K. P. Hewagamage and G. K. A. Dias.
University of Colombo School of Computing,
35, Reid Avenue, Colombo 7, Sri Lanka.
E-mail: saminda@webmail.cmb.ac.lk, {ddk ,gnw ,kph ,gkd }@ucsc.cmb.ac.lk
Abstract experience to a more flexible and learner-centric
experience [3]. As a result of the increasing availability
Use of video clips for e-learning is very limited due to the of e-learning, the market for training in workplace
high usage of band width. The ability to select and readiness and problem-solving is growing rapidly.
retrieve relevant video clips using semantics addresses Establishing virtual universities and colleges and
this problem. This paper presents a Profile based Feature digital libraries, developing online courses and content
Identification system for multimedia database systems are all important activities to support e-learning. They
which is designed to support the use of video clips for e- enable remote access to a vast volume of educational
learning. This system is capable of storing educational material anytime for e-learners, who can then spend their
video clips with their semantics and retrieving required limited time on understanding and processing material on
video clip segments efficiently on their semantics. The their own pace. A large volume of digital documents
system creates profiles of presenters appearing in the that can be used for e-learning are currently available on
video clips based on their facial features and uses these the internet in different forms such as text files, image
profiles to partition similar video clips into logical files, voice clips, video clips, question databases etc.. In
meaningful segments. The face recognition algorithm addition, the distance learning systems augment this
used by the system is based on the Principal Components volume of digital video documents on the internet every
Analysis (PCA) approach. However PCA algorithm has day.
been modified to cope with the face recognition in video Integration of heterogeneous data as content for e-
key frames. Several improvements have been proposed to learning applications is crucial, since the amount and
increase the face recognition rate and the overall versatility of processable information is the key to a
performance of the system. successful system. Multimedia database systems can be
used to organize and manage heterogeneous multimedia
e-learning content [8]. At the same time, the large
1. Introduction amount of visual information, carried by video
documents requires efficient and effective indexing and
In today's rapidly changing electronic world (e-world) the searching tools. The development of standards for video
key to maintain the appropriate momentum in encoding such as the XML-based MPEG-7 standard
organizations and academic environments is knowledge. introduced by the moving pictures expert group (MPEG)
Therefore, continuous, convenient and economical access coupled with the increased power of computing made
to training material assumes the highest priority for the that content-based manipulation of digital video
ambitious individual or organization. This requirement is information feasible [5].
met by electronic learning (e-learning). E-learning is one Another important aspect that determines the success
of the fastest growing areas of the high technology sector of a e-learning system is how efficiently the system uses
today and is a highly cost-effective and adaptable medium the available bandwidth. One solution to this problem is
for education and training. to provide facilities for the user to browse and select
E-learning offers potentially universal access to what he actually required before delivering the material.
content, regardless of location, and it can transform This can be done by categorizing and clustering various
education and training from a passive consumption
74
2. types of educational materials by using ontologies and However in our system we use profiles to annotate
indices. semantics to video clips automatically. The system also
In this paper our focus is on video based educational provides features to extend the metadata associated with
material where presenters deliver educational content. We profiles later at any time as they become available. The
employ a set of tools developed by us to segment video annotated metadata is saved in a XML database. We use
clips semantically into shots by using low level features. XML databases for metadata because it allows both
Then we identify those segments where presenters appear multimedia educational objects and metadata to be stored
and extract the relevant information in key face frames. and handled uniformly by using the same techniques.
These information are then encoded and compared with a The remainder of this paper is organized as follows.
database of similarly encoded images. The feature The system architecture is shown in Section two. Section
information in video frames of a face is represented as an three reviews a number of techniques related to our work.
eigenvector which is considered as a profile of a Section four explains the technique for segmenting face
particular person [17]. These profiles are then used to regions and describes the use of PCA (Principle
construct an index over the video clips to support efficient Component Analysis) for our work. The implementation
retrieval of video shots. of the system is shown in Section five and the results
Once the profiles for the presenters are created a obtained are shown in Section six. Finally, in Section
semi-automatic semantic annotation process is used to seven gives our conclusions and address the future work
annotate meta-data with the video shots. Majority of based on this project.
automatic metadata authorization procedures reported in
the literature are based on the video’s physical features
such as color, motion, or brightness data [22, 23,].
Course Materials
Ontology
Keyword
Keyword Organizer
Extractor
Video Shot Feature Profile Index
Channel Identification Extractor Creator
Video Filter
Clips
Audio Semantic Profile
Channel Segmentation Creation
Processor
Links
Meta-Data
Object Profiles Database
Multimedia (XML DB)
Object server
Query Processor
Figure 1: System Architecture
2. System Architecture The system stores this educational material in a
multimedia object server. The keyword extractor extracts
The overall architecture of our system is shown in Figure keywords from the main course materials. The keyword
1. The main components of our system are the keyword organizer assists the construction of an ontology in a
extractor, keyword organizer, Feature extractor, Profile database out of the keyword generated by the keyword
creator and the query processor. extractor. The feature extractor extracts audio and video
Various types of course materials such as course notes, features from the video clips and the profile creator
PowerPoint presentations, quizzes, past examination creates profiles of presenters from the information
papers and video clips are the main input to this system. generated by the feature extractor. These profiles are
75
3. then used to create indices on the video clips. Finally the
query processor process enables the end users to browse
and retrieve educational material stored in the object
server by using the ontology and the indices. Our system extracts the following types of
descriptors for each of the key-frames.
2.1 Video Segmentation
1. Color histogram
Video segmentation can be done either manually or 2. Edge histogram
automatically. Manual segmentation is usually time-
consuming but more accurate. Many approaches to 2.2 Profile Creation.
automate segmentation of video sequences have been
proposed in the past [21, 22, 23]. Earlier approaches The system initially uses a set of video clips from a video
exploited mostly the motion information in order to library to compute the eigenvectors of presenters [17].
extract moving objects from a scene [15]. However, most An eigenvector computed for a presenter in this way can
of the contemporary techniques have merged motion be thought of as a point in the possible eigenspace. Due
information with information obtained from edge to various reasons the eigenvectors compute for the same
extraction and/or texture analysis to increase the accuracy presenter by using different shots may result in multiple
[22, 23]. non equal eigenvectors. These eigenvectors can be
In our system a video is analyzed by segmenting it into thought of as a set of features that together characterize
shots, selecting key-frames, and extracting audio-visual the variation between face images. In such cases a single
descriptors from the shots (See Figure 2). This allows the eigenvector is created by correlating the individual
video to be searched at the shot-level using content-based eigenvectors created for that presenter by considering the
retrieval approaches. fact that faces possess similar structure (eye, nose and
Our approach initially uses a semi-automatic method mouth position, etc). One of the main reasons for using
on a training data set to construct profiles of presenters. eigenfaces for our research is that it needs a lower
These profiles are subsequently used to automatically dimensional space to describe faces.
assign semantics to the video shots. We have primarily
investigated models that apply broadly to video content, 2.3 Audio Segmentation.
such as presenter vs. slide show, change of presenter,
change of speaker and change of lecture etc. While the
In addition to automatic analysis and modeling of the
models allow the video content to be annotated
features of the video content, we also investigated the
automatically using this small vocabulary, the integration
use of speech indexing to combine our approach for
of the different search methods together like content-
video retrieval.
based and model-based allow more effective indexing and
In the audio stream, initial segmentation was carried
retrieval (See Figure 1).
out through the use of the Bayesian Information
Input Video Criterion (BIC) [16]. The technique used in this system
is based on the variable window scheme proposed by
… …. Tritschler & Gopinath [16]. The Expectation
Shot Detection
Maximization algorithm was applied for the training of
the Gaussian Mixture Models (GMM) for the known
speakers [12]. Mel Frequency Cepstral Coefficients
(MFCC) features were extracted from the “unknown”
audio and tested against the GMM speaker model. The
… ….
output of this procedure is a likelihood value for the
Key Frame Extraction
speaker in the given audio stream.
2.4 Multimedia Object Server
All the multimedia objects are indexed and saved on a
XML database. We are using Apache Xindice 1.0 as our
multimedia object server and MPEG-7 Description
Feature Extraction Schemes schemas to store the multimedia metadata [8].
The Description Schemes (DS) provide a standardized
Figure 2 Segmentation of video clips way of describing in XML the important concepts related
76
4. to audio-visual content description and content Using neural networks for face recognition is another
management in order to facilitate searching, indexing, popular approach. Steve Lawrence has developed a face
filtering, and access. A relational database is used to store recognition system based on Self Organizing Maps
the profiles and catalogues. (SOMs) and Convolutional Neural Networks (CNN)
[10]. Their system consists of an SOM fed into a
3. Face Detection and Recognition Convolutional Neural network. The problem with the
SOM is that it arbitrarily divides input space into a set of
In the field of multimedia, the focus of research has been classes of which the designer has no control or
not just detection but also identification of faces, people knowledge. Another problem with the neural networks
or some specific objects in video images or video find is a result of their inability to deal with the high
footages. A face recognition system can be thought of as dimensionality of the problem. For an example, when we
being comprised of two stages: consider a image of size 128 * 128 pixels requires a
neural net with 16,384 input neurons for processing.
1. Face Segmentation Furthermore, to train such a neural network, and ensure
2. Face Recognition robust performance requires an extremely large training
set (much bigger than 16,384). This is often not possible
The first step of any face processing system is in real-world applications where only one or two images
detecting the locations in images where faces are present. of an individual may be available.
However, face detection from a single image is a Proposed in 1991 by Turk and Pentland, this was a
challenging task because of variability in scale, location, successful system for automatic recognition of human
orientation (up-right, rotated), and pose [1]. In general faces [17]. This method can be classified as appearance-
single face detection methods are classified into the based methods, which uses the whole face region as the
following four categories: raw input to a recognition system. The objective of an
appearance-based face recognition algorithm is
1. Knowledge-based methods essentially to create low-dimensional representations of
2. Feature invariant approaches face images to perform recognition. In contrast,
3. Template matching methods geometric feature-based methods attempt to distinguish
4. Appearance-based methods between faces by comparing properties and relations
between facial features, such as eyes, mouth, nose and
However these methods have overlap category chin. As a consequence, success of these methods
boundaries. The algorithms of the first category are depends on the feature extraction and measurement
simple. In general, algorithms of this type are used to process.
detect faces in real time when the volume of data involved
is small [4]. Most of the time, the algorithms of the 4. Profile Construction Algorithm
second and fourth categories are implemented on
expensive workstations dedicated to image processing and Motivated by the work of Paul Viola and Michael Jones
employee real time processing [6]. [18], we use a new image representation called an
There are many approaches for face recognition integral image that allows for very fast feature evaluation.
ranging from the Principal Component Analysis (PCA) We use a set of features which are reminiscent of Haar
approach (also known as eigenfaces) [17], Elastic Graph Basis functions. In order to compute these features very
Matching (EGM) [9], Artificial Neural Networks [10, 14], rapidly at many scales we used the integral image
to Hidden Markov Models (HMM) [2]. All these systems representation for key frames. The integral image is
differ in terms of the feature extraction procedures and/or computed from an image using a few operations per
the classification techniques used. pixel. Once computed, any one of these Haar-like
Michael C. Lincoln and Adrian F. Clark of the features are computed at any scale or location very fast
University of Essex have proposed a scheme for [6].
independent face identification in video sequences [11]. We use AdaBoost to construct a classifier by selecting
In their research an “unwrapped” texture map is a small number of important features [19]. Feature
constructed from a video sequence using a texture-from- selection is achieved through a simple modification of
motion approach. A drawback with unwrapped texture the AdaBoost procedure: the weak learner is constrained
map is the recognition will be only comparable to the best so that each weak classifier returned depends on only a
front-face-only frames. Unlike this technique, eigenfaces single feature. As a result each stage of the boosting
are robust against noise and poor lighting conditions. Also process, which selects a new weak classifier, can be
eigenfaces are relatively insensitive to small variation in viewed as a feature selection process.
scale, rotation and expression.
77
5. The complete face detection cascade has 32 classifiers, for each person, with some variations in expression and
which total over 80,000 operations. Nevertheless the in the lighting. Some of the eigenfaces that are stored in
cascade structure results in extremely rapid average our database is shown in Figure 4.
detection times.
Figure 3 shows some face detection samples from
different video segments. Operating on 352 x 288 pixel
image frames, it takes less then 1 second to detect faces.
So the approach is extremely efficient and fast. After
detecting the faces, the face segments are passed in to the
face recognition system based on PCA.
Figure 4: Eigenfaces from profile database
Figure3: Detected face samples
There is an average eigenface for each class as well
Our method of face recognition is based on profiles, and this is considered as a profile of person.
which is created by using principle component analysis If there is M total eigenvectors, the average matrix
(PCA) [17]. Among the best possible known approaches has to be calculated and then subtracted from the original
for face recognition, Principal Component Analysis
(PCA) has been object of much effort. In PCA, the faces and the result stored in the variable :
recognition system is based on the representation of the
face images using the so called eigenfaces. In the
eigenface representation, every training image is
considered a vector of pixel gray values (i.e. the training
images are rearranged using row ordering).
An eigenvector of a matrix is a vector such that, if
multiplied with the matrix, the result is always an integer Then the covariance matrix C is calculated
multiple of that vector. This integer value is the according to,
corresponding eigenvalue of the eigenvector. This
relationship is described by the equation below.
A×u=λ×u
Then the eigenvectors (eigenfaces) and the
corresponding eigenvalues are calculated. The
Where u is an eigenvector of the matrix A (n × n) and
eigenvectors (eigenfaces) are normalized so that they are
λ is the corresponding eigenvalue. unit vectors of length 1. From M eigenvectors
Eigenvectors possess following properties:
(eigenfaces), only M’ are chosen, which have the highest
eigenvalues. The higher the eigenvalue, the more
• They can be determined only for square characteristic features of a face does the particular
matrices
eigenvector describe. Eigenfaces with low eigenvalues
• There are n eigenvectors (and corresponding are omitted, as they explain only a small part of
eigenvalues) in an n × n matrix. characteristic features of the faces [17]. After M’
• All eigenvectors are perpendicular, i.e. at right eigenfaces are determined, the ”training” phase of the
angle with each other. algorithm is finished.
There is a problem with the algorithm described in
The system functions by projecting face images onto a equation 3. The covariance matrix C has a
feature space that spans the significant variations among
dimensionality of × , so we would have
known face images. The significant features are known as
eigenfaces and eigenvalues. For a 128 × 128 image
"eigenfaces" because they are the eigenvectors (principal
that means that one must compute a 16,384 × 16,384
components) of the set of faces. Face images are collected
matrix and calculate 16,384 eigenfaces. Computationally,
into sets. Every set (or class) includes a number of images
78
6. this is not very efficient as most of those eigenfaces are Each of the most frontal faces is normalized into a
not useful for our task. 128 x 128 image using the eye positions, and then
converted to a point in the 16-dimensional eigenspace.
5. Implementation
Figure 5 shows the structure of an educational video clip
segments in which several presenters appearing. As
shown in the diagram, face features and voice features
are extracted from the video clips by analyzing the audio
Where L is a M × M matrix, v are M eigenvectors of L and video channels separately. The system also employs
and u are eigenfaces. video-caption recognition to obtain face-voice-name
The covariance matrix C is calculated using the association if captions are available on the video clips,
formula C = . The advantage of this method is that otherwise the user is expected to enter this meta-data
one has to evaluate only M numbers and not . Usually, manually. In many cases, a video caption is attached to a
face and usually represents a presenter’s name. So video-
M << as only a few principal components
caption recognition provides rich information for face-
(eigenfaces) is be relevant. The amount of calculations to
voice-name association.
be performed is reduced from the number of pixels
Given the extracted faces voices and names, the
( × ) to the number of images in the training set (M).
indexing system combines the corresponding data
We use only a subset of M eigenfaces, the M’ eigenfaces
together and creates the required indices to support
with the largest eigenvalues.
information retrieval. Finally the query processor
The process of classification of a new (unknown) face
responds to different types of user queries by using these
to one of the classes (known faces) proceeds in two
indices and the multimedia object server as shown.
steps. First, the new image is transformed into its
eigenface components. The resulting weights w form the
weight vector :
6. Evaluation
The Techniques that we have explained in this paper
have been evaluated by developing a prototype system.
A collection of video clips already used to deliver
educational content to one of our external degree
program over the TV is used as the input to our system.
The Euclidean distance between two weight vectors From this collection we first created a medium size
provides a measure of similarity between the database with profiles of 10 people. For each person we
corresponding images i and j. If the Euclidean distance have chosen 10 face video frames with different imaging
between and other faces exceeds on average some conditions. After the construction of this initial profile
database, a random sample of 65 key frames were
threshold value , we assume that is no face at all.
selected from our video collection and tested with our
also allows one to construct ”clusters” of faces system. A small number presented poor imaging
such that similar faces are assigned to one cluster. conditions which our algorithms were not designed to
Let an arbitrary instance x be described by the feature accommodate. These conditions included very dark
vector lighting different camera angles and head orientation
more that 30 degrees.
Our system achieves a recognition rate of 92% when
Where denotes the value of the r th attribute of we tested on 10 face classes (see Figure 6) and it dropped
instance x. Then the distance between two instances to 70% when we added another 10 face classes to our
and is defined to be database. Recognition results of up to 80.5% were
obtained for 20 face classes that contain straight looking
faces.
79
7. !" " "
#"
" #
% &
#"
" $ ! #
Figure 5: Implementation
small. We have proposed a simple and efficient
technique to detect and recognize human faces in a video
sequence but are two major challenges: the illumination
and pose problems. Pose discrimination is not difficult
but accurate pose estimation is hard to acquire.
We tested the performance of our implementation by
varying the number of face classes for different number
of eigenfaces. We observe that the algorithm is sensitive
to the number of face classes. The recognition rate
decreases when we increase the number of classes,
because in eigenspace some face classes can overlap for
some faces that have similar facial features.
In order to increase the recognition rate, methods that
decrease the number of classes should be explored. One
of these methods can be constructing a hierarchical tree
structure. If we consider the top level nodes as main face
classes, each node must have a small number of child
Figure 6: Results for 20 face classes nodes which contains sub classes with attributes of facial
features extracted in different poses. This method will
7. Conclusion and Future Work improve the pose problem in face recognition for some
extent.
Recognition of faces from a video sequence is still one of Nevertheless, as far as face recognition in video
the most challenging problems in face recognition sequences is concerned, much work still remains to be
because video is of low quality and the frame images are done.
80
8. [12] Liu, M., Chang, E., and Dai, B. (2002). Hierarchical
8. Acknowledgement. Gaussian Mixture Model for Speaker Verification.
Proceedings International Conference on Spoken
This work is supported by the Japan International Language Processing.
Cooperation Agency (JICA) and Asian Development
Bank (ADB). The authors would like to thank the all the [13] Lorente, L., and Torres, L. (1998). Face Recognition of
reviewers for insightful comments. The authors also Video Sequences in a MPEG-7 Context Using a Global
acknowledge each individual appearing in our face Eigan Approach. International Workshop on Very Low
database. Bit-rate Video Coding, Urbana, Illinois.
[14] Palanivel, S., Venkatesh B. S., and Yegnanarayana, B.
References (2003). Real Time Face Recognition System Using
Autoassociative Neural Network Models. IEEE
[1] Adini, Y., Moses, Y., and Ullman, S. (1993). Face International Conference on Acoustics, Speech and Signal
Recognition: the Problem of compensating for Changes in Processing, Hong Kong, 833-836.
Illumination Direction. Technical Report CS93-21,
Mathematics & Computer Science, Weizmann Institute of [15] Park, S., Park, J., and Aggarwal, J. K. (2003). Video
Science. Retrieval of Human Interaction Using Model-Based
motion Tracking and Multi layer Finite State Automata.
[2] Bicego, M., Castellani, U., and Murino, V. (2003). Using Image and Video Retrieval, Second International
Hidden Markov Models and Wavelets for face recognition. Conference, CIVR 2003, Urbana-Champaign, IL, USA,
Image Analysis and Processing, 2003.Proceedings of 12th 394-403.
International Conference,52-56.
[16] Tritschler, A., Gopinath, R. A. (1999) Improved Speaker
[3] E-Learning: Putting a World-Class Education at the Segmentation and Segments Clustering Using the
Fingertips of all Children. (2000) The National Educational Bayesian Information Criterion. Sixth European
Technology Plan, U.S Department of Education, December. Conference on Speech Communication and Technology.
[4] Fr¨oba, B., Ernst, A., and K¨ublbeck, C. (1998). Real-Time [17] Turk, M., and Pentland, A. (1991). Eigenfaces for
Face Detection. Department of Applied Electronics, Recognition. Journal of Cognitive Neuroscience, 71-86.
Fraunhofer Institute for Integrated Circuits, Germany.
[18] Viola, P., and Jones, M. (2001). Fast and Robust
[5] ISO/IEC JTC1/SC29/WG11 Coding of Moving Pictures and Classification using Asymmetric AdaBoost and a Detector
Audio. International Organization for Standardization. Cascade. Neural Information Processing Systems.
http://zola.byu.edu/vad/byu2.pdf.
[19] Viola. P., and Jones, M. (2001). Robust Real-time Object
[6] Kawato, S., and Ohya, J. (2000). Two-step approach for real- Detection. Second International Workshop on Statistical
time eye tracking with a new filtering technique. International and computational theories of vision Canada.
Conference on Systems, Man and Cybernetics.
[20] Wang H. and Chang, S. (1996). Automatic face region
[7] Kobla, V., Doermann, D., Lin, K., and Faloutsos, C. (1997). detection in MPEG video sequences. Conference
Compressed domain video indexing techniques using DCT Electronic Imaging and Multimedia Systems, part of
and motion vector information in MPEG video. Storage and SPIE’s Photonics, China.
Retrieval for Image and Video Databases (SPIE), 200-211.
[21] Yeo, B., Liu, B. (1995). Rapid scene analysis on
[8] Kosch, H. (2000). MPEG-7 and Multimedia Database compressed video. IEEE Transactions on Circuits &
Systems. SIGMOD Records, ACM Press, 34–39. Systems for Video Technology, 533-44.
[9] Lades, M., Vorbriiuggen, J. C., Buhmann, J., Lange, J., [22] Zabih, R., Miller J., and Mai, K. (1995). A feature-based
Malsburg, C., Wiiurtz, R. P., and Konen, W. (1993). algorithm for detecting and classifying scene breaks. Proc.
Distortion Invariant Object Recognition in the Dynamic Link ACM Multimedia, 189-200.
Architecture. IEEE Transactions on Computers, 300-311.
[23] Zhang, H., Kankanhalli A., and Smoliar, W. (1993).
[10] Lawrence, S., Giles, L. C., Tsoi, A. C. and Back, A. D. Automatic partitioning of full-motion video”, Multimedia
(1997). Face Recognition: A Convolutional Neural Network Systems, 10-28.
Approach. IEEE Transactions on Neural Networks, 98 -113.
[11] Lincoln, M. C., and Clark, A. F. (2001). Pose-Independent
Face Identification from Video Sequences. VASE Laboratory,
University of Essex Colchester CO4 3SQ, UK, AVBPA, 14-19
81