Self-Organising Maps as a Relevance Feedback Technique in Content-Based Image Retrieval

Abstract

Self-Organising Maps (SOMs) can be used in implementing a powerful relevance feedback mechanism for Content-Based Image Retrieval (CBIR). This paper introduces the PicSOM CBIR system, and describes the use of SOMs as a relevance feedback technique in it. The technique is based on the SOM's inherent property of topology-preserving mapping from a high-dimensional feature space to a two-dimensional grid of artificial neurons. On this grid similar images are mapped in nearby locations. As image similarity must, in unannotated databases, be based on low-level visual features, the similarity of images is dependent on the feature extraction scheme used. Therefore, in PicSOM there exists a separate tree-structured SOM for each different feature type. The incorporation of the relevance feedback and the combination of the outputs from the SOMs are performed as two successive processing steps. The proposed relevance feedback technique is described, analysed qualitatively, and visualised in the paper. Also, its performance is compared with a reference method.

A Novel Shape Feature for Image Classification and Retrieval

“…The classification performance of different features was tested with the KNearest Neighbor leave-one-out cross-validation and the built-in CBIR analysis system of a PicSOM [10], a content-based image retrieval system developed at the Laboratory of Computer and Information Science at Helsinki University of Technology. The main features of PicSOM are efficient indexing based on treestructured self-organizing maps and adaptive querying using relevance feedback.…”

Interactive retrieval of video using pre-computed shot-shot similarities

“…We envision the proposed data organisation as a directed graph, but this is not unique. Stemming from a different paradigm, in the feature vectors into tree-structured self-organising maps [16], relevance for images is determined by their distance on the map to the user-judged examples.…”

Image Theft Detection with Self-Organising Maps

“…In the past, self-organising maps [5] have been used on a number of image processing and retrieval tasks such as content-based image retrieval [8,9,10,11,12], automatic image annotation [14], colour-based image browsing [15] and others. The use of tree-structured variants, which allow fast logarithmic search [6,7], therefore presents itself for copyright theft detection.…”