Intelligent Television (Intelligent-TV)

To develop an intelligent, computerised, and automatic TV content watch and recording system, which is capable of: (i) learning the users’ preference and interests in watching TV programmes; (ii) automatically extracting knowledge from the TV content via combination of text mining, visual semantics analysis, and audio semantics analysis; (iii) selectively recording all the TV content which may be preferred and desired by consumers; (iv) searching, retrieving and managing the recorded TV content via a simple remote control set; (v) automatically upgrading itself by deleting the obsolete content and adding new content.

With TV channels being increased at an explosive speed including Internet TV, mobile TV, analogue TV, digital TV and satellite TV etc. users are being spoiled and confused with too many options. As the time passes by, valuable information and interesting programmes are being missed even without their knowledge. On the broadcasting side, it remains a challenge how to produce the right TV content and ensure that the content is watched by the desired audience. On the other hand, the cost of digital storage is constantly decreasing, providing a viable case for new recording products with significantly large storage spaces. The proposed project will develop an intelligent TV watching system to help consumers to automatically keep all the possibly interesting TV programmes for their leisure time to browse, to search, and to selectively watch or retrieve for other purposes. The project will also provide new technologies to help the content producers to ensure that the content produced will reach the right audience, and thus the efficiency and effectiveness of their content production will be significantly improved. The research impact can be highlighted in two ways: (i) new technology innovation and new product invention for intelligent digital recoding (DVD recorders) to automatically watch one TV channel 24 hours a day and 365 days a year, or more expensive products to watch two or more TV channels. As a result, users will have significant digital TV content to browse through, to select, and to enjoy without worrying about missing anything. (ii) new research areas and algorithms on digital content knowledge extraction, knowledge based digital video content analysis and processing, leading to further research on human-like digital content consumption, perception, and intelligent management. Possible workpackages are outlined as follows:

WP-1: Project management and specification design;
WP-2: Adding semantics at production side (annotation and summarization, embedded watermarking);
WP-3: Semantics analysis of TV content (video semantics and audio semantics analysis);
WP-4: Capture of users’ habits, preferences, and interests in watching TVs;
WP-5: Hybrid search engine for management of recorded digital content (viewing text summarization or annotation, semantics based search and retrieval, and thumbnail browsing in compressed domain etc.);
WP-6: System integration, validation, dissemination and exploitation, a prototype system is to be built up to illustrate the project concept and commercial value via: (i) monitoring one TV channel; (ii) content recording capacity of around 500GB; (iii) users validation, evaluation and demonstration with such a system by using a remote control set to retrieve programmes, set automatic recording mode, and show compatibility with existing DVD recording products or technologies;

 

new television technologies;
semantics-based TV content processing and analysis;
text mining, search and match;
audio-visual knowledge extraction and analysis;
consumer electronics and system integration;
television broadcasters.

(i) research partners in (a) text mining, search and match; (b) audio content processing and analysis;

(ii) consumer electronics companies for system integration and platform build-up to illustrate the intelligent television concept;

(iii) television broadcasting companies for (a) digital content annotation, authoring, and summarization; (b) technology validation by producing and broadcasting new concept TV content.

(i) manual text annotation and summary of TV content from production side;

(ii) text based intelligent TV content match, search, and selection;

(iii) audio based intelligent TV content match, search, and selection;

(iv) system integration and platform build up for the project (A 500GB-based DVD recording of one TV channel is preferred, which is capable of interfacing with computerised software for intelligent TV content processing);

(v) TV broadcasting and TV content production arrangement for dissemination and exploitation of the project concept.

(i) research partners specialising in text-based mining, semantics extraction, search and match;

(ii) research partners specialising in audio-based semantics extraction, search, match and content selection towards intelligent video analysis;

(iii) consumer electronics company;

(iv) TV broadcasting company.

No