About this role
These positions are part of the Joint Project “Understanding Multimedia Content (UMC)” of the Cognitive Computing research line (http://ict.fbk.eu/cognitive-computing) at FBK. The UMC project is developed in collaboration between the Natural Language Processing research unit (hlt.fbk.eu) the Technologies for Vision research unit
(tev.fbk.eu) and the Data and Knowledge Management research unit (dkm.fbk.eu).
MULTIMEDIA INFORMATION EXTRACTION DRIVEN BY BACKGROUND KNOWLEDGE: (http://ict.unitn.it/application/project_specific_grants#A3)
This phd has the objective of extracting events from commented videos exploiting background knowledge available in the semantic web. This phd should develop a holistic approach, where the process of extracting information from the video, and from the associated text are integrated and can affect each other at any stage. This implies that video stream and textual stream are considered as a whole information space and their interpretations are not independent. Furthermore, video-text interpretation should not happen in the knowledge vacuum, but it should exploit the existing large amount of background knowledge available in the semantic web under the form of ontologies and RDF data. Nowadays–in contrast with the early years of AI when knowledge acquisition was a bottleneck–large amount of common-sense knowledge is available in the semantic web, but it cannot be easily exploited by the state-of-the-art approaches to video and text analysis. The thesis should investigate on how to extend and adapt algorithms for video and text analysis in order to inject background knowledge. The thesis, to reach it’s objective, should combine techniques in machine learning–for processing low level data–with automated reasoning–to manage with high level semantic knowledge.
VISION FOR MULTIMEDIA UNDERSTANDING:
(http://ict.unitn.it/application/project_specific_grants#A4)
Multimedia content analysis more and more relies on advanced machine learning to capture the enormous richness of multi-modal sources (commented videos, images with captions, etc.). At the other side, domain specific knowledge is often available to leverage the content analysis task, but effectively encoding it into machine learning (dow to the development of task-specific feature representations) is still an open research issue. The goal of this PhD is to progress on the computer vision side of the problem, to go beyond a mono-modal approach where supervisions for learning are provided explicitly. Instead, we will investigate how structured (background knowledge) and semi-structured data (e.g. text captions and descriptions) can be used to provide implicit supervision to enrich the task-specific visual learning capabilities.
Company:
FBK
Qualifications:
Candidates are required to have basic or advanced skills in one or more of the following topics: artificial intelligence, machine learning, computer vision, natural language processing and knowledge representation techniques.
Specific requirements:
Duration: 3 years
Educational level:
Master Degree
How to apply:
Please mention NLP People as a source when applying