Grupo de Tratamiento de Imágenes

 

 

 

 

 

News and Events 

Retrieving Events in Life Logging

The CLEF Initiative (Conference and Labs of the Evaluation Forum) is a self-organized body whose main mission is to promote research, innovation, and development of information access systems. Since 2000 the CLEF has played a leading role in stimulating investigation and research in a wide range of key areas in the information retrieval domain, becoming well-known in the international information retrieval (IR) community.

ImageCLEF 2018 is an evaluation campaign that is being organized as part of the CLEF initiative labs in 2018. The campaign offers several research tasks that welcome participation from around the world. ImageCLEF2018lifeLog is one of the main tasks of ImageCLEF2018, with the main goal is to make use of lifelogging data and explore the possibilities that come with it. The task offered two subtasks: Lifelog Moment Retrieval (LMRT) and Activities of Daily Living Understanding (ADLT).

ImageCLEFlifelog2018 received in total 41 runs: 29 (21 official, 8 additional) for LMRT and 12 (8 official, 4 additional) for ADLT, from 7 teams from Brunei, Taiwan, Vietnam, Greece-Spain, Tunisia, Romania, and a multi-nation team from Ireland, Italy, Austria, and Norway. The AILabGTi team submitted eight runs and achieved the best score F1@10 at 0.545, the highest official score, for the LMRT subtask.

They proudly invited AILabGTi team to present our study (the original letter here):

Kavallieratou, E., Del Blanco, C.R., Cuevas, C., Garcia, N.: “Retrieving Events in Life Logging”

as an oral presentation at the CLEF 2018 Conference, hosted by the University of Avignon, France, from 10 to 14 September 2018.

Abstract: This paper describes our contribution for the Lifelog Moment Retrieval (LMRT) challenge of ImageCLEF Lifelog2018. Lifelogging has a tremendous potential in many applications. However, the wide range of possible moment events along with the lack of fully annotated databases make this task very challenging. This work proposes an interactive and weakly supervised learning approach that can dramatically reduce the time to retrieve any kind of events in huge databases. Impressive results have been obtained in the referred challenge, reaching the first rank.

You can download it here.