CBIR-ANR: A content-based image retrieval with accuracy noise reduction
Gabriel da Silva Vieira, Afonso U. Fonseca, Fabrízzio Soares
Abstract
Due to the expansion of multimedia data leveraged by social networks and smartphone devices, large sets of information are available and increasing daily. In this context, information retrieval is crucial to open new opportunities to individuals, governments, and businesses. Therefore, we present the CBIR-ANR software in which the content-based image retrieval (CBIR) is followed by an accuracy noise reduction (ANR) strategy that adjusts query responses and increases assertiveness in image retrieval. Also, the software combines three low-level features to form a 187-dimensional feature vector, which is size efficient for large-scale data sets and competitive with related work.
Topics & Concepts
Computer scienceContent-based image retrievalImage retrievalContext (archaeology)SoftwareInformation retrievalNoise (video)Feature (linguistics)Image (mathematics)Artificial intelligenceProgramming languageLinguisticsBiologyPaleontologyPhilosophyImage Retrieval and Classification TechniquesAdvanced Image and Video Retrieval TechniquesVideo Analysis and Summarization