Alignment between a technical paper and presentation sheets using a hidden markov model

Hayama, Tessai; Nanba, Hidetsugu; Kunifuji, Susumu

doi:10.1109/amt.2005.1505278

Cited by 22 publications

(15 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In our future work, we are planning to develop slide applications using the structure data extracted by this technique and the slide information processing technique [2] that we have developed.…”

Section: Discussionmentioning

confidence: 99%

“…The likelihood of the candidate attribute contains not only the likelihood degree of the candidate attribute but also the unlikelihood degree of the other attributes. Thus, the likelihood of the candidate attribute is defined and its value (Li Attri) is given by equations (1) and (2). Here, attri, Attri V al (attri) and M axScore (attri) indicate an attribute, its scores assigned in step 1 and the max score for each attribute 3 , respectively.…”

Section: ) Assigning the Likelihood Of Each Attribute Within Each Obmentioning

confidence: 99%

See 1 more Smart Citation

Structure Extraction from Presentation Slide Information

Hayama

Nanba

Kunifuji

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Electronic presentations are used in numerous scenarios, such as lectures and meetings. In recent years, the widespread use of electronic presentations means that presentation slide data is increasing as one of industry's most important information resources. Therefore, it is necessary to develop a practical usage method for the reutilisation of the data on slides. An approach to achieve this is to focus on visual structure information within a slide, because visual structure information is one of the most valuable, easy to understand methods for humans. However, since visual structure information is not explicitly defined in the slide data itself, computers have difficulty comprehending structure information directly. In this paper, we propose a method of extracting structure information from slide information. The proposed method is composed of two steps: organising objects within the slide as units, such as title, body text, figure and table, and structuring the units as a hierarchy tree based on a top-down approach.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: ) Assigning the Likelihood Of Each Attribute Within Each Obmentioning

confidence: 99%

Structure Extraction from Presentation Slide Information

Hayama

Nanba

Kunifuji

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Similar to (Hayama et al, 2005;Kan, 2007), we have previously proposed how presentation slides are used to compile reference summaries automatically (Zhang et al, 2008). The motivations behind this procedure are:…”

Section: Active Learning For Tagging Reference Summary and Summarizationmentioning

confidence: 99%

Active learning of extractive reference summaries for lecture speech summarization

Zhang

Fung

2009

Proceedings of the 2nd Workshop on Building and Using Comparable Corpora From Parallel to Non-Parallel Corpora - BUCC '09

View full text Add to dashboard Cite

We propose using active learning for tagging extractive reference summary of lecture speech. The training process of feature-based summarization model usually requires a large amount of training data with high-quality reference summaries. Human production of such summaries is tedious, and since inter-labeler agreement is low, very unreliable. Active learning helps assuage this problem by automatically selecting a small amount of unlabeled documents for humans to hand correct. Our method chooses the unlabeled documents according to the similarity score between the document and the comparable resource-PowerPoint slides. After manual correction, the selected documents are returned to the training pool. Summarization results show an increasing learning curve of ROUGE-L F-measure, from 0.44 to 0.514, consistently higher than that of using randomly chosen training samples.

show abstract

“…Despite the critical importance of speech in presentations, much research that supports presentations focuses on slide creation [1][2][3]. To help presenters create additional explanations in speech, Kojiri et al developed a system that analyzes and visualizes the relations among slides that audiences can recognize [4].…”

Section: Introductionmentioning

confidence: 99%

Effective Presentation Speech Support System for Representing Emphasis-Intention

Kojiri

Kaji

2015

Systems

View full text Add to dashboard Cite

A research presentation integrates slides and speech. If these two aspects do not represent the same intention, the presentation will probably fail to effectively explain the presenter's intention. This paper focuses on the representation of the critical contents in a presentation. In an effective speech, the speaker adds more intonation and stress to emphasize the importance of the slide contents. Audiences recognize that important contents are those that are explained in a stronger voice or that are said after a short pause. However, in ineffective speeches, such voice effects do not always correspond to the important contents that are indicated by slides. On slides, the important contents are represented by levels of text indentation and size, color, and animation. This research develops a presentation speech support system that estimates important contents from slides and voices that might be recognized by audiences and extracts numerical differences. In addition, the system provides comments and feedback to improve speeches.

show abstract

Alignment between a technical paper and presentation sheets using a hidden markov model

Cited by 22 publications

References 3 publications

Structure Extraction from Presentation Slide Information

Structure Extraction from Presentation Slide Information

Active learning of extractive reference summaries for lecture speech summarization

Effective Presentation Speech Support System for Representing Emphasis-Intention

Contact Info

Product

Resources

About