GPGPU implementation of evolutionary algorithm for images clustering
D Konieczny, M Marcinkowski… - Advanced Methods for …, 2013 - Springer
D Konieczny, M Marcinkowski, PB Myszkowski
Advanced Methods for Computational Collective Intelligence, 2013•SpringerWe propose an evolutionary algorithm (EA) usage to image clustering applied to Document
Search Engine (DSE). Each document is described by its visual content (including images),
preprocessed and clustered by EA. Next, such clusters are core of DSE. However, number of
documents and attached images make EA ineffective in such task. Using the natural issue of
EA we propose GPGPU (General Purpose Graphic Processing Unit) implementation. The
paper describes our proposition, research and results gained in experiments.
Search Engine (DSE). Each document is described by its visual content (including images),
preprocessed and clustered by EA. Next, such clusters are core of DSE. However, number of
documents and attached images make EA ineffective in such task. Using the natural issue of
EA we propose GPGPU (General Purpose Graphic Processing Unit) implementation. The
paper describes our proposition, research and results gained in experiments.
Abstract
We propose an evolutionary algorithm (EA) usage to image clustering applied to Document Search Engine (DSE). Each document is described by its visual content (including images), preprocessed and clustered by EA. Next, such clusters are core of DSE. However, number of documents and attached images make EA ineffective in such task. Using the natural issue of EA we propose GPGPU (General Purpose Graphic Processing Unit) implementation. The paper describes our proposition, research and results gained in experiments.
Springer
Showing the best result for this search. See all results