This paper summarizes an ongoing work in multimedia processing aimed at the automated archiving and selective retrieval of textual, pictorial and auditory information contained in video programs. Video processing performs the task of representing the visual information using a small subset of the video frames. Linguistic processing refines the closed caption text, generates table of contents, a...

Video structuring is the process of extracting temporal structural information of video sequences and is a crucial step in video content analysis especially for sports videos. It involves detecting temporal boundaries, identifying meaningful segments of a video and then building a compact representation of video content. Therefore, in this paper, we propose a novel mechanism to automatically pa...

With constant advances in information technology, more and more information is available and users’ information needs are becoming more diverse. Most conventional information systems only attempt to provide information that meets users’ specific interests. In contrast, we are working on ways of discovering information from the viewpoints of both interest and necessity. For example, we are tryin...

Videotext recognition is challenging due to low resolution, diverse fonts/styles, and cluttered background. Past methods enhanced recognition by using multiple frame averaging, image interpolation and lexicon correction, but recognition using multi-modality language models has not been explored. In this paper, we present a formal Bayesian framework for videotext recognition by combining multipl...

We have designed and constructed a mechanism for using caption text from broadcast television programs to analyze their content. In this paper, we describe the method by which captions are captured and translated from the raw video signal into text on the ViewStation. We also describe our Caption Parser, which analyzes the text and extracts information about the content of broadcast television ...

We present a novel data selection method for lightly supervised training of acoustic model, which exploits a large amount of data with closed caption texts but not faithful transcripts. In the proposed scheme, a sequence of the closed caption text and that of the ASR hypothesis by the baseline system are aligned. Then, a set of dedicated classifiers is designed and trained to select the correct...

The paper addresses a scheme of lightly supervised training of an acoustic model, which exploits a large amount of data with closed caption texts but not faithful transcripts. In the proposed scheme, a sequence of the closed caption text and that of the ASR hypothesis by the baseline system are aligned. Then, a set of dedicated classifiers is designed and trained to select the correct one among...

This paper presents a method of automatically creating hypermedia documents from conventional transcriptions of television programs. Using parallel text alignment techniques, the temporal information derived from the closed caption signal is exploited to convert the transcription into a synchronized text stream. Given this text stream, we can create links between the transcription and the image...

This paper describes cognitive characteristics of timing difference for closed captions superimposed onto TV news programs. It was reported that timing delays for superimposing disrupts hearing impaired people's enjoyment and intelligibility of TV, but nobody has yet investigated the permissible limit for timing difference. This study presents subjects' permissible and preferable limits of the ...

The increased use of video documents for multimedia-based applications has created a demand for strong video database support, including efficient methods for browsing and retrieving video data. Most solutions to video browsing and retrieval of video data rely on visual information only, ignoring the rich source of the accompanying audio signal and texts. Speech is the significant information t...