scispace - formally typeset
Search or ask a question
Topic

Feature hashing

About: Feature hashing is a research topic. Over the lifetime, 993 publications have been published within this topic receiving 51462 citations.


Papers
More filters
Journal ArticleDOI
TL;DR: This work analyzes twitter data using support vector machine algorithm to classify tweets into positive, negative and neutral sentiments, and finds the relationship between feature hash bit size and the accuracy and precision of the model that is generated.
Abstract: Sentiment Analysis is a way of considering and grouping of opinions or views expressed in a text. In this age when social media technologies are generating vast amounts of data in the form of tweets, Facebook comments, blog posts, and Instagram comments, sentiment analysis of these usergenerated data provides very useful feedback. Since it is undisputable facts that twitter sentiment analysis has become an effective way in determining public sentiment about a certain topic product or issue. Thus, a lot of research have been ongoing in recent years to build efficient models for sentiment classification accuracy and precision. In this work, we analyse twitter data using support vector machine algorithm to classify tweets into positive, negative and neutral sentiments. This research try to find the relationship between feature hash bit size and the accuracy and precision of the model that is generated. We measure the effect of varying the feature has bit size on the accuracy and precision of the model. The research showed that as the feature hash bit size increases at a certain point the accuracy and precision value started decreasing with increase in the feature hash bit size. General Terms Hadoop, Data Processing, Machine learning
Journal ArticleDOI
TL;DR: Wang et al. as mentioned in this paper proposed a deep hashing model called adaptive hash code balancing (AHCB), which introduces a balanced binary method to maximize the hash value entropy so that the generated hash has better clustering.
Abstract: ABSTRACT Remote sensing image retrieval (RSIR) refers to finding images from an image database that contain the same instance as the query image, which is an essential task in remote sensing applications. Traditional depth-based hashing algorithms usually convert the image library into a hash matrix with a specified number of digits. On the one hand, the quality of hash matrices generated by traditional methods is low and cannot guarantee good clustering between pictures of the same class. On the other hand, the ability to extract features using backbone networks must be improved. This paper proposes a deep hashing model called adaptive hash code balancing (AHCB) to solve two existing problems. The model introduces a balanced binary method to maximize the hash value entropy so that the generated hash has better clustering. Graph convolutional networks(GCNs) are introduced to automatically detect relevant data points in the graph, perform back-propagation, and propagate the updated network feedback to the feature extraction layer to improve the ability to extract features. It enables the model to learn the intrinsic data structure between remote sensing images. Experimental results on three public datasets show that the proposed method outperforms the current state-of-the-art deep hashing remote sensing image retrieval algorithms by a large margin.
Journal ArticleDOI
TL;DR: Wang et al. as discussed by the authors proposed Weakly Supervised Hashing with Reconstructive Cross-modal Attention (WSHRCA) to learn compact visual-semantic representation with more reliable supervision for retrieval task.
Abstract: On many popular social websites, images are usually associated with some meta-data such as textual tags, which involve semantic information relevant to the image, and can be used to supervise the representation learning for image retrieval. However, these user-provided tags are usually polluted by noise, therefore the main challenge lies in mining the potential useful information from those noisy tags. Many previous works simply treat different tags equally to generate supervision, which will inevitably distract the network learning. To this end, we propose a new framework, termed as Weakly Supervised Hashing with Reconstructive Cross-modal Attention (WSHRCA), to learn compact visual-semantic representation with more reliable supervision for retrieval task. Specifically, for each image-tag pair, the weak supervision from tags are refined by cross-modal attention, which takes image feature as query to aggregate the most content-relevant tags. Therefore, tags with relevant content will be more prominent while noisy tags will be suppressed, which provides more accurate supervisory information. To improve the effectiveness of hash learning, the image embedding in WSHRCA is reconstructed from hash code, which is further optimized by cross-modal constraint and explicitly improves hash learning. The experiments on two widely-used datasets demonstrate the effectiveness of our proposed method for weakly-supervised image retrieval. The code is available at https://github.com/duyc168/weakly-supervised-hashing.
Journal ArticleDOI
TL;DR: Zhang et al. as discussed by the authors proposed a new image hashing scheme based on saliency map and sparse model, which combines a visual attention model called Itti model and the matrix of color vector angle (CVA).
Abstract: Abstract Image hashing is an effective technology for extensive image applications, such as retrieval, authentication and copy detection. This paper designs a new image hashing scheme based on saliency map and sparse model. The major contributions are twofold. The first contribution is the construction of a weighted image representation by combining a visual attention model called Itti model and the matrix of color vector angle (CVA). Since the Itti model can efficiently detect saliency map and CVA fully captures color information of image, they contribute to a visually robust and discriminative image representation. The second contribution is the hash extraction from the weighted image representation via sparse model. A classical sparse model called robust principal component analysis is exploited to decompose the weighted image representation into a low-rank component and a sparse component. As the low-rank component can describe intrinsic structure of image, hash calculation with low-rank component can achieve good discrimination. The efficiencies of the proposed scheme are validated by extensive experiments with open databases. The results demonstrate that the proposed scheme is superior to some state-of-the-art schemes in terms of classification performance between robustness and discrimination.

Network Information
Related Topics (5)
Feature extraction
111.8K papers, 2.1M citations
84% related
Convolutional neural network
74.7K papers, 2M citations
84% related
Feature (computer vision)
128.2K papers, 1.7M citations
84% related
Deep learning
79.8K papers, 2.1M citations
83% related
Support vector machine
73.6K papers, 1.7M citations
83% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
202333
202289
202111
202016
201916
201838