The third Workshop on Automated Information Extraction in Media Production (AIEMPro10) aims at fostering exchange of ideas and of practices between leading experts in research and leading actors in the media community, in order to catalyze the migration towards new ways of producing media content, aided by large scale introduction of tools for automated multimedia analysis and understanding. Furthermore, the workshop helps researchers in better understanding what are some real-life key requirements which would enable their scientific developments come into wider adoption.

In this paper, we describe the latest specifications of the Metadata Production Framework (MPF) and its reference software with which the basic functionality of MPF can be tested over networks. MPF was originally designed as an efficient environment for generating metadata for TV programs from the viewpoint of the broadcaster, but recent developments indicate that the domains where MPF can be applied have expanded to include even ordinary households. This trend suggests that the metadata market might be much bigger in the near future, and we believe MPF can become an infrastructure standard in these circumstances.

This paper describes our attempt to generate quizzes automatically from news programs, as novel method for content creation. Of the many types of possible quizzes we focus on the format of a multiple-choice quiz with accompanying images. To begin, we worked on formulating the engineering problem of quiz generation - selection of images suitable for a quiz, selection of text that explains the image, and the generation of alternative choices that are similar but different by selecting from subtitles. All the approaches were designed based on principal object in the image and text. As the suitable image for the quiz, an image which has a clear subject - we defined it as spotlight-image - is introduced. As for generation of the similar but different choices, categorizing the subject into person or other can make the correspondence between images and sentence much clear, and then appropriate choices are generated. By introducing these techniques, the accuracy has risen (35.7% up) and the effectiveness of our proposed method were confirmed.

This paper proposes a method for generating image-based quizzes from news video achieves. Although there are many types of quizzes, in this work we focus on matching quizzes in which an image is to be matched to one of several choices that are statements. The key to making a successful quiz of this type is to extract choice statements that are similar but nonidentical to the true statement regarding the image. For image selection, we need to select images that are suitable for image-based quizzes. In this paper, we report some preliminary work and highlight some interesting issues in making quizzes. We also propose a method for automatically generating quizzes using feature-based clustering analyses and assess the outcomes of trials actually using our method.

This paper proposes the Metadata Production Framework (MPF) as a common platform for generating content-based metadata. A lot of research on extracting useful information from audiovisual content has been conducted. Recently it has been necessary to integrate these research studies to get much higher semantic metadata. The aim of MPF is to provide an environment where we can easily make these types of processes, including multimodal integration. We have also released a Metadata Editor where a user can generate content-based metadata both manually and by using automatic metadata extraction modules based on MPF.

This paper describes a method to generate metadata for TV programs in real-time by utilizing acoustic and speech data in live broadcasting. Various styles of watching TV programs can be provided by using metadata related to the content of the program. The acoustic data to be processed in our case is crowd noise in a football (soccer) stadium, and the speech data is an announcer's voice. The crowd noise is closely related to not only spectator emotions but also their attention and expectations. In other words, a part in which the crowd noise rises corresponds to an important event in the game. Because the crowd noise conveys no further information about what happened in the scene, the announcer's voice, after speech-to-text conversion, is processed to extract further meaning. By combining these two processes of identifying and extracting, content-based segment metadata is generated automatically. This method was applied to generating metadata for six professional football games, by which its effectiveness was verified.