Intelligent Systems Lab Amsterdam

Image semantics

Image retrieval departs from the idea that an understanding of the image content comes from analysis invariant features and understanding its meaning by means of machine learning. We investigate ways to close the semantic gap in content-based image retrieval as well as ways to track arbitrary objects over time. In the end, we aim at a full verbal description of the image content by researching the parallels of natural language and image interpretation.

Participants

• Jan van Gemert
• Jan-Mark Geusebroek
• Arjan Gijsenij
• Theo Gevers
• Cees Snoek
• Nicu Sebe
• Arnold Smeulders (contact)
• Jasper Uijlings
• Cor Veenman
• Marcel Worring