Image Retrieval using Long-Term Semantic Learning

Matthieu Cord (1) , Philippe-Henri Gosselin (2)

1 ETIS - UMR 8051 - Equipes Traitement de l'Information et Systèmes
2 MIDI - Multimedia Indexation and Data Integration

The automatic computation of features for content-based image re- trieval still has difficulties to represent the concepts the user has in mind. Whenever an additional learning strategy (such as relevance feedback) can improve the results of the search, the system perfor- mances still depend on the representation of the image collection. We introduce in this paper a supervised optimization of a set of fea- ture vectors. According to an incomplete set of partial labels, the method improves the representation of the image collection, even if the size, the number, and the structure of the concepts are unknown. Experiments have been carried out on a large generalist database in order to validate our approach.

Domaines: Machine Learning [stat.ML]

Matthieu Cord, Philippe-Henri Gosselin. Image Retrieval using Long-Term Semantic Learning. IEEE International Conference on Image Processing, Oct 2006, France. pp.1. ⟨hal-00520307⟩