Proyecto

TEC2007-65400 (SemanticVideo): SemanticVideo: Estrategias de análisis en vivo de secuencias de vídeo para indexación semántica y acceso personalizado (2008-2010)

El objetivo de este proyecto es investigar en el diseño y desarrollo de algoritmos de análisis de
secuencias de vídeo para su aplicación en adaptación de contenido audiovisual. La principal
innovación propuesta parte de la hipótesis de que la realimentación entre etapas de análisis y la
inclusión sistemática y formalizada de información contextual en dichas etapas es una vía para
mejorar sustancialmente la calidad de los resultados obtenidos y para independizar los algoritmos de
análisis de su contexto de aplicación, dos de los principales problemas que presenta el estado del
arte en este ámbito. El objetivo es lograr un alto grado de fiabilidad en la extracción de
información semántica de secuencias de video, información que supone el punto de partida para el
segundo campo de innovación propuesto: la adaptación semántica de contenido
Financiado por: Plan Nacional de I+D+I
Participantes: Universidad Autonoma de Madrid (VPU)
Investigador responsable: Martinez Sanchez, Jose María

Enlace Web: TEC2007-65400 (SemanticVideo)

Publicaciones Externas:

Revistas

• Fernando Barreiro, Jose M. Martinez, Víctor Valdés : "Visual Tools for ROI Montage in an Image2Video Application", IEEE Transactions on Circuits and Systems for Video Technology (En prensa)


• Luis Herranz, Jose M. Martinez : "An integrated approach to summarization and adaptation using H.264/MPEG-4 SVC", Signal Processing: Image Communication


• Luis Herranz, Jose M. Martinez : "On the use of hierarchical prediction structures for efficient summary generation of H.264/AVC bitstreams", Signal Processing: Image Communication


• Luis Herranz, Jose M. Martinez : "A framework for scalable summarization of video", IEEE Transactions on Circuits and Systems for Video Technology


• Juan C. SanMiguel, Jose M. Martinez : "On the use of feedback strategies in the detection of events for video surveillance", IET Computer Vision

Congresos

• Victor Fernandez-Carbajales, Miguel A. García, Jose M. Martinez : "Robust People Detection by Fusion of Evidence from Multiple Methods", 9th International Workshop on Image Analysis for Multimedia Interactive Services 2008


• Juan C. SanMiguel, Jesús Bescós, Jose M. Martinez, Alvaro García : "DiVA: a Distributed Video Analysis framework applied to video-surveillance systems", 9th International Workshop on Image Analysis for Multimedia Interactive Services 2008


• Juan C. SanMiguel, Jose M. Martinez : "Robust unattended and stolen object detection by fusing simple algorithms", 5th IEEE Internacional Conference on Advanced Video and Signal based Surveillance 2008


• Marcos Escudero, Fabrizio Tiburzi, Jesús Bescós : "MPEG video object segmentation under camera motion and multimodal backgrounds", 2008 IEEE International Conference on Image Processing


• Alvaro Bayona, Juan C. SanMiguel, Jose M. Martinez : "Comparative evaluation of stationary foreground object detection algorithms based on background subtraction techniques", 6th IEEE Internacional Conference on Advanced Video and Signal based Surveillance 2009


• Juan C. SanMiguel, Jose M. Martinez, Alvaro García : "An Ontology for Event Detection and its Application in Surveillance Video", 6th IEEE Internacional Conference on Advanced Video and Signal based Surveillance 2009


• Fabrizio Tiburzi, Marcos Escudero, Jesús Bescós, Jose M. Martinez : "A ground-truth for Motion-based Video-object Segmentation", 2008 IEEE International Conference on Image Processing


• Fernando López, Jose M. Martinez : "Automatic adaptation decision making using an image to video adaptation tool in the MPEG-21 framework", 10th International Workshop on Image Analysis for Multimedia Interactive Services


• Fernando López, Dietmar Jannach, Jose M. Martinez, Christian Timmerer, Hermman Hellwagner : "Multimedia Adaptation Decisions Modelled as Non-Deterministic Operations", 9th International Workshop on Image Analysis for Multimedia Interactive Services 2008


• Víctor Valdés, Jose M. Martinez : "On Video Abstraction Systems' Architectures and Modelling", Third International Conference on Semantics and Digital Media Technologies


• Jose M. Martinez, Víctor Valdés : "Binary Tree Based On-Line Video Summarization", 16th ACM International Conference on Multimedia


• Fernando Barreiro, Jose M. Martinez, Víctor Valdés : "Virtual Camera Tools for an Image2Video Application", 9th International Workshop on Image Analysis for Multimedia Interactive Services 2008


• Juan C. SanMiguel, Jose M. Martinez : "Shadow Detection in video surveillance by maximizing the agreement between independent detectors", 2009 IEEE International Conference on Image Processing


• Luis Herranz, Jose M. Martinez : "An efficient summarization algorithm based on clustering and bitstream extraction", 2009 IEEE International Conference on Multimedia and Expo


• Fernando López, Jose M. Martinez, Narciso García : "CAIN-21: An extensible and metadata-driven multimedia adaptation engine in the MPEG-21 framework", 4th International Conference on Semantic and Digital Media Technologies (En prensa)


• Juan C. SanMiguel, Jose M. Martinez : "Dynamic video surveillance systems guided by domain ontologies", 3rd International Conference on Imaging for Crime Detection and Prevention 2009


• Luis Herranz, Jose M. Martinez : "On the advantages of the use of bitstream extraction for video summary generation", 16th International Conference on MultiMedia Modeling


• Marcos Escudero, Jesús Bescós : "Mean-Shift Combination of Illumination-Invariant Features For Video Object Segmentation", The 6th European Conference on Visual Media Production


• Marcos Escudero, Jesús Bescós : "Region-based video object segmentation robust to illumination", 11th International Workshop on Image Analysis for Multimedia Interactive Services


• Juan C. SanMiguel, Andrea Cavallaro, Jose M. Martinez : "Evaluation of on-line quality estimators for object tracking", 2010 IEEE International Conference on Image Processing


• Alvaro Bayona, Juan C. SanMiguel, Jose M. Martinez : "Stationary foreground detection using background subtraction and temporal difference in video surveillance", 2010 IEEE International Conference on Image Processing


• Juan C. SanMiguel, Jose M. Martinez : "On the evaluation of background subtraction algorithms without ground-truth", 7th IEEE International Conference on Advanced Video and Signal based Surveillance 2010


• Alvaro Garcia-Martin, Jose M. Martinez : "Robust Real Time Moving People Detection in Surveillance Scenarios", 7th IEEE International Conference on Advanced Video and Signal based Surveillance 2010


• Marcos Escudero, Jesús Bescós : "A robust framework for region based video object segmentation", 2010 IEEE International Conference on Image Processing (En prensa)


• Juan C. SanMiguel, Marcos Escudero, Jose M. Martinez, Jesús Bescós : "Real-time single-view video event recognition in controlled environments", 9th International Workshop on Content-based Multimedia indexing


• Alvaro Garcia-Martin, Alexander Hauptmann, Jose M. Martinez : "People Detection Based on Appearance and Motion Models", 8th IEEE International Conference on Advanced Video and Signal based Surveillance 2011


• Alvaro García, Jesús Bescós : "Video Object Segmentation Based on Feedback Schemes Guided by a Low-Level Scene Ontology", 10th International Conference on advanced Concepts for Intelligent Vision Systems

Publicaciones Internas:

Tesis

• Advances in semantic-guided and feedback-based approaches for video analysis, SanMiguel Avedillo, Juan Carlos, Octubre 2011, (Tutor: Martinez Sanchez, Jose María)


• Integración realimentada de esquemas de coherencia espacial y temporal de regiones en la segmentación de objetos por modelado de fondo en secuencias de vídeo, Escudero Viñolo, Marcos, (Tutor: Bescós Cano, Jesús)

Trabajos de fin de master

• People detection algorithms based on appearance and motion information, Garcia-Martin, Alvaro, Noviembre 2009, (Tutor: Martinez Sanchez, Jose María)


• Ontology-guided semantic video description using feedback between signal processing stages, SanMiguel Avedillo, Juan Carlos, Noviembre 2008, (Tutor: Martinez Sanchez, Jose María)

Proyectos de fin de carrera

• Segmentación espacial de secuencias MPEG en el dominio comprimido, Escudero Viñolo, Marcos, Febrero 2008, (Tutor: Tiburzi Paramio, Fabrizio) , (Ponente: Bescós Cano, Jesús)

Informes Técnicos

• Rafael Martin, Juan C. SanMiguel, Jose M. Martinez: Detección de interacciones persona-objeto en secuencias de vídeo-seguridad no basada en seguimiento de objetos


• Alvaro Garcia-Martin, Juan C. SanMiguel, Victor Fernandez-Carbajales, Miguel A. García, Jose M. Martinez: Event detection for the TRECVID 2009 video surveillance dataset: the VPULab-UAM contribution


• Rafael Martin, Juan C. SanMiguel, Jose M. Martinez: Informe sobre evaluación de algoritmos de detección de eventos en sistemas de vídeo-seguridad


• Juan C. SanMiguel, Javier Molina, Jesús Bescós, Jose M. Martinez: Informe sobre diseño de ontologías de dominio