Logiques de descriptions pour l'analyse structurelle de film (original) (raw)

Analyse automatique de film-Des séquences d'images aux séquences d'actions.

2009

Résumé Dans le contexte de l'annotation de documents filmiques assistée par ordinateur, nous posons le problème de l'analyse filmique automatique, et identifions 3 problèmes de base: classification de plans, regroupement temporel et pilotage d'algorithmes d'extraction. Ces trois processus mettent en œuvre des connaissances provenant de plusieurs domaines d'expertise: documentalistes, spécialistes d'analyse du signal et professionnels de l'audiovisuel.

A survey on description and modeling of audiovisual documents

Multimedia Tools and Applications, 2020

The number of audiovisual documents available on the web is exponentially increasing due to the rise of the number of videos produced every day. The recent progress in audiovisual documents field has made it possible to popularize the exchange of these documents in many domains. More generally, the interest in the indexing potential of audiovisual documents has significantly increased in different disciplines, namely films, sports events, etc. Within this framework, several research studies focused on implementing this indexation based on the segmentation of the audiovisual document in fragments. This segmentation was brought by the appropriate descriptions. Although the indexing process seems essential, the way of exploiting and searching audiovisual documents remains unsatisfactory. Indeed, annotations based on generic descriptions (title, creator, publisher, etc.) are insufficient to describe the content of the audiovisual documents. With the proliferation of audiovisual documents and the mentioned indexing limits, the question that should be answered is: "What is the relevant information of the audiovisual content?". In this paper, we present a survey to characterize the description and the modeling of audiovisual documents. We classify the existing description methods into three categories: a low-level description, a documentary description and a semantic description. The main objective of this study is to propose an approach that helps describe and organize the content of an audiovisual document so as to conduct a better inquiry.

Analysis of Animation Movies: Segmentation, Abstraction and Annotation

2005

In acest articol este propus un sistem de analiză a filmelor de animaţie. Prima etapă de analiză constă in detecţia tranziţiilor video. Pentru a depaşii problemele ridcate de specificitatea filmelor de animaţie, sunt propuse o serie de metode imbunătăţite de detecţie a tranziţiilor "cuts" şi "fades". De asemenea, folosind un detector de "flash" modificat, este propusă o metodă de detecţie a unui efect specific filmelor de animaţie, denumit "short color change". In final este prezentată o metodă automată de rezumare si anotare a conţinutului video. Sistemul a fost testat folosind mai multe filme de animaţie.

Extraction of information of audio-visual contents

Proceedings of The 4th International Conference on the Foundations of Information Science

In this article we show how it is possible to use Channel Theory (Barwise and Seligman, 1997) for modeling the process of information extraction realized by audiences of audiovisual contents. To do this, we rely on the concepts proposed by Channel Theory and, especially, its treatment of representational systems. We then show how the information that an agent is capable of extracting from the content depends on the number of channels he is able to establish between the content and the set of classifications he is able to discriminate. The agent can endeavor the extraction of information through these channels from the totality of content; however, we discuss the advantages of extracting from its constituents in order to obtain a greater number of informational items that represent it. After showing how the extraction process is endeavored for each channel, we propose a method of representation of all the informative values an agent can obtain from a content using a matrix constituted by the channels the agent is able to establish on the content (source classifications), and the ones he can understand as individual (destination classifications). We finally show how this representation allows reflecting the evolution of the informative items through the evolution of audiovisual content.

Audio-based description and structuring of videos

Enabling a rapid on-the-fly view of the content of a movie requires segmenting the movie and describing the segments in a user-compatible manner. The difficulty resides in extracting relevant semantic information from the audiovisual signal, both for the segmentation and the description. We introduce in this paper audio scenes and chapters in movies and present an algorithm for automatically segmenting a video based on the audio stream only. Audio scenes and chapters are defined as the equivalent of shots and scenes in the visual domain. A tree-like audio-based structure of a video is proposed. A chapter is then classified into different chapter categories. The automatic solution to audio scene and chapter segmentation and classification is evaluated on manually segmented and classified videos

CLAVIS - Un systeme de raisonnement temporel pour la classification de sequences audiovisuelles

Recherche d'Information Assistee par Ordinateur, 2000

In the context of video indexing, we present the Clavis system in which typical video sequences of television programs are represented by templates. Templates are terminological constraint networks in which video segments coming from automatic analysis tools are represented in a description logic formalism. Templates allow to express complex classes of video sequences with temporal constraints associated to a regular expression operator. Recognizing occurrences of a template in a video program is a plan recognition problem for which efficient methods have been implemented in a constraint satisfaction problem framework. The paper describes the system and illustrates its use with several experiments that were done in the context of the DiVAN european project.

Content Identification for Audiovisual Archives

International Association of Sound and Audiovisual Archives (IASA) Journal

The collective value of a network of archives follows the pattern established by Metcalfe's Law, with the network's value increasing at a greater rate than the simple sum of its individual members. (

AdA Filmontology – Levels, Types, Values

AdA Toolkit, 2021

by Thomas Scherer, Jasper Stratil, Yvonne Pfeilschifter, Rebecca Zorko, Anton Buzal, João Pedro Prado, Henning Agt-Rickauer, Christian Hentschel, Harald Sack, Matthias Grotkopp, and Jan-Hendrik Bakels. Version 1.0 (July 2021) https://www.ada.cinepoetics.fu-berlin.de/en/ada-toolkit/index.html The AdA Filmontology is a systematic vocabulary and data model of film-analytical terms and concepts for fine-grained semantic video annotations. The vocabulary was developed in close collaboration between film scholars and computer scientists as a tool for digital film studies. Its goal is to provide a standardized and systematic basis for the joint annotation of audio-visual corpora to enable comparable, systematic film analyses. The vocabulary is grounded in a methodological film-analytical consensus and is made available as a machine-readable OWL ontology to publish annotations as Linked Open Data for the exchange and comparison of analysis data.