This talk presents an overview of Visual information retrieval (VIR), an active and vibrant research area, which attempts at providing means for organizing, indexing, annotating, and retrieving visual information (images and videos) from large, unstructured repositories.
It gives special emphasis to the process of searching and retrieving images using a visual query, usually known as Content-Based Image Retrieval (CBIR). This field of research has experienced significant scientific and technological advances during the past 25 years, which have led to today's commercial solutions capable of being presented with an image and answering questions such as: What is this?, Where can I purchase it?, Who painted this?, or Can you find me more images like this?
Covered topics include:
- Principles of Visual Information Retrieval (VIR)
- Examples of contemporary VIR systems (research prototypes and commercial solutions)
- Designing and evaluating a VIR system
- Datasets, challenges, and benchmarks for VIR
- Medical image retrieval
- Mobile visual search
- Case studies and success stories
- Future directions in VIR.