Deep Learning for Image Retrieval: What Works and What Doesn't

Wang, Huafeng; Cai, Yehe; Zhang, Yanxiang; Pan, Haixia; Liu, Weifeng; Han, Hongbin

doi:10.1109/icdmw.2015.121

Cited by 39 publications

(19 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The research undertaken by [13] reveals the use of deep learning by linking different designs and metric choices to improve the efficiency of the image retrieval system. In addition to these, researchers [14] and [15] had applied deep learning for image retrieval. More applications of deep learning can be seen in the work by [16] which used deep learning for fast cover song retrieval and the work by [17] applied deep learning for multimedia retrieval.…”

Section: Related Workmentioning

confidence: 99%

An Embodied Conversational Agent using Retrieval-Based Model and Deep Learning

Leong*¹,

Goh²,

Kumar³

2019

IJITEE

View full text Add to dashboard Cite

In accordance to the Fourth Industrial Revolution, the emergence of the digital transformation alongside with the collaboration between networked machines and human beings in decision-making has attracted much attention globally. The utilization of conversational agent to convey information on behalf of human have gained popularity and been used extensively. As a result, this research intends to examine and to innovate the current corporate website by providing the capabilities to deliver instantaneously replies and reliable information through the integration of a conversational agent via deep learning, comparable to communicating with the competent customer service consultant. This study is reliant on the Artificial Intelligence (AI) to offer natural language processing (NLP) by presenting retrieval-based model and Deep Learning to enable the conversation agent to make smarter and better decision in generating reliable and up-to-date responses. Furthermore, this research intends to automate the process of minimizing the need for human interference, specifically the botmaster to perform the knowledge maintenance manually.

show abstract

Section: Related Workmentioning

confidence: 99%

An Embodied Conversational Agent using Retrieval-Based Model and Deep Learning

Leong*¹,

Goh²,

Kumar³

2019

IJITEE

View full text Add to dashboard Cite

show abstract

“…The experiments have shown effective CBIR system with the combination of features than using individual features. Many recent works [8,9,10,11,12] have used the convolutional neural network for extraction of features from the images and store them. These methods have shown superior performances than the earlier methods of feature extraction.…”

Section: Related Workmentioning

confidence: 99%

Content based Image Retrieval System using Combination of Color and Shape Features, and Siamese Neural Network

2019

IJITEE

View full text Add to dashboard Cite

With an advent of technologya huge collection of digital images is formed as repositories on world wide web (WWW). The task of searching for similar images in the repository is difficult. In this paper, retrieval of similar images from www is demonstrated with the help of combination of image features as color and shape and then using Siamese neural network which is constructed to the requirement as a novel approach. Here, one-shot learning technique is used to test the Siamese Neural Network model for retrieval performance. Various experiments are conducted with both the methods and results obtained are tabulated. The performance of the system is evaluated with precision parameter and which is found to be high.Also, relative study is made with existing works.

show abstract

“…CNN, specifically designed to manage the variability of 2D shapes, has proven to have superior performance to , which is able to learn from the feature maps and classify multimodal images with different variability using a common flow Unlike the previous methods, the binarization approaches [5,6] require pair wise entries for learning the binary code, the representation of the features has the best performance provided by CNN, the ability to generalize the extracted characteristics, the relationship between dimensional reduction and loss of precision in CBIR, the best distance measurement technique in CBIR and the advantage of coding techniques to improve the efficiency of CBIR, [5] have proposed to present a binarization approach of the characteristics for a better CBIR efficiency. More precisely, the binarization has reduced the use of the 31/32 space of the original data.…”

Section: Related Workmentioning

confidence: 99%

An Advanced Relevance Feedback Method to Improve Performance of CBIR using Convolutional Neural Network and Comprehensive Values

Kumar¹,

Saradhi²

2019

IJEAT

View full text Add to dashboard Cite

Content-Based Image Retrieval (CBIR) is extensively used technique for image retrieval from large image databases. However, users are not satisfied with the conventional image retrieval techniques. In addition, the advent of web development and transmission networks, the number of images available to users continues to increase. Therefore, a permanent and considerable digital image production in many areas takes place. Quick access to the similar images of a given query image from this extensive collection of images pose great challenges and require proficient techniques. From query by image to retrieval of relevant images, CBIR has key phases such as feature extraction, similarity measurement, and retrieval of relevant images. However, extracting the features of the images is one of the important steps. Recently Convolutional Neural Network (CNN) shows good results in the field of computer vision due to the ability of feature extraction from the images. Alex Net is a classical Deep CNN for image feature extraction. We have modified the Alex Net Architecture with a few changes and proposed a novel framework to improve its ability for feature extraction and for similarity measurement. The proposal approach optimizes Alex Net in the aspect of pooling layer. In particular, average pooling is replaced by max-avg pooling and the non-linear activation function Maxout is used after every Convolution layer for better feature extraction. This paper introduces CNN for features extraction from images in CBIR system and also presents Euclidean distance along with the Comprehensive Values for better results. The proposed framework goes beyond image retrieval, including the large-scale database. The performance of the proposed work is evaluated using precision. The proposed work show better results than existing works.

show abstract

Deep Learning for Image Retrieval: What Works and What Doesn't

Cited by 39 publications

References 6 publications

An Embodied Conversational Agent using Retrieval-Based Model and Deep Learning

An Embodied Conversational Agent using Retrieval-Based Model and Deep Learning

Content based Image Retrieval System using Combination of Color and Shape Features, and Siamese Neural Network

An Advanced Relevance Feedback Method to Improve Performance of CBIR using Convolutional Neural Network and Comprehensive Values

Contact Info

Product

Resources

About