Seminar from Sadok Mansouri, ATER at LIUM

 

Date: 15/11/2024
Time: 11h00
Place: IC2, Boardroom
Speaker: Sadok Mansouri
 
 

Information Extraction and Analysis from News videos

 

Information extraction from videos is an important research topic in content-based video indexing and retrieval. Indeed, the visual text present in news videos typically provides rich semantic information, such as person identities, location names, and event types, which serve as a high-level index. However, variations in text due to differences in size, style, orientation, and alignment, as well as the linguistic complexity of same text, particularly in Arabic language , make the problem of automatic information extraction extremely challenging.

In this talk, I will introduce our implementation solution called SISAVIN (Semantic Indexing System for Arabic News Videos), which includes four subtasks: text detection and recognition, text classification, entity recognition, and event extraction. The SISAVIN framework is based on recent methods in computer vision and natural language processing (NLP).