Open Access Open Access  Restricted Access Subscription or Fee Access

Multiclass Content-based Visual Information Retrieval using Multi Layer Perceptron and Convolution Neural Network: A Comparative Study

Shamik Tiwari

Abstract


Content-based visual information extraction is an area, which utilizes the computer vision methods for searching and indexing the images in image library more competently than manual annotations. Due to the exponentially increase in image database volume, there is a demand of efficient retrieval system. Deep learning-based machine vision systems have the ability to handle large amount of data efficiently than ordinary machine vision systems. This paper proposed a content-based image search system, which utilizes convolution neural network based deep learning model for multi-class content-based image search. The results show the superiority of convolution neural network model on multi-layer perceptron model.

Keywords: Image informatics, deep learning, multi-layer perceptron, convolution neural network

Cite this Article: Shamik Tiwari. Multi-class Content-based Visual Information Retrieval using Multi-Layer Perceptron and Convolution Neural Network: A Comparative Study. Journal of Artificial Intelligence Research & Advances. 2020; 7(1): 1–9p.


Full Text:

PDF


DOI: http://dx.doi.org/10.37591%2Fjoaira.v7i1.1880

Refbacks

  • There are currently no refbacks.