Filters








6,123 Hits in 4.3 sec

Region-DH: Region-based Deep Hashing for Multi-Instance Aware Image Retrieval

Franck Romuald Fotso Mtope, Bo Wei
2020 2020 International Joint Conference on Neural Networks (IJCNN)  
Region-DH can flexibly be used with existing deep neural networks or more complex object detectors for image hashing.  ...  We design a unified deep neural network that simultaneously localizes and recognizes objects while learning the hash functions for binary codes.  ...  deep Convolutional Neural Networks (referred to as deep hashing networks) to analyze labeled images, extract usable patterns and further create an improved feature representation for hash functions [  ... 
doi:10.1109/ijcnn48605.2020.9207485 dblp:conf/ijcnn/MtopeW20 fatcat:zvyrgns65vfabgyui6sy2rxiwa

A Survey on Deep Hashing Methods

Xiao Luo, Haixin Wang, Daqing Wu, Chong Chen, Minghua Deng, Jianqiang Huang, Xian-Sheng Hua
2022 ACM Transactions on Knowledge Discovery from Data  
Hashing is one of the most widely used methods for its computational and storage efficiency. With the development of deep learning, deep hashing methods show more advantages than traditional methods.  ...  In this survey, we detailedly investigate current deep hashing algorithms including deep supervised hashing and deep unsupervised hashing.  ...  ACKNOWLEDGMENTS This work was supported by the National Key Research and Development Program of China (2021YFF1200902) and the National Natural Science Foundation of China (31871342).  ... 
doi:10.1145/3532624 fatcat:7lxtu2qzvvhrpnjngefli2mvca

Learning to Authenticate with Deep Multibiometric Hashing and Neural Network Decoding [article]

Veeru Talreja, Sobhan Soleymani, Matthew C. Valenti, Nasser M. Nasrabadi
2019 arXiv   pre-print
The MDHND consists of two separate modules: a multimodal deep hashing (MDH) module, which is used for feature-level fusion and binarization of multiple biometrics, and a neural network decoder (NND) module  ...  In this paper, we propose a novel multimodal deep hashing neural decoder (MDHND) architecture, which integrates a deep hashing framework with a neural network decoder (NND) to create an effective multibiometric  ...  In this framework, we leveraged a neural network based decoder to refine the codes generated by the deep hashing network to improve the authentication performance.  ... 
arXiv:1902.04149v3 fatcat:mjhplnruuvf6dczz4llgf6fywm

A Survey on Deep Hashing Methods [article]

Xiao Luo, Haixin Wang, Daqing Wu, Chong Chen, Minghua Deng, Jianqiang Huang, Xian-Sheng Hua
2022 arXiv   pre-print
Hashing is one of the most widely used methods for its computational and storage efficiency. With the development of deep learning, deep hashing methods show more advantages than traditional methods.  ...  In this survey, we detailedly investigate current deep hashing algorithms including deep supervised hashing and deep unsupervised hashing.  ...  ACKNOWLEDGMENTS This work was supported by the National Key Research and Development Program of China (2021YFF1200902) and the National Natural Science Foundation of China (31871342).  ... 
arXiv:2003.03369v5 fatcat:m2iu3htilvgztkcazw3cyk6iqe

Deep Residual Hashing [article]

Sailesh Conjeti, Abhijit Guha Roy, Amin Katouzian, Nassir Navab
2016 arXiv   pre-print
The DRH model constitutes four key elements: (1) a sub-network with multiple stacked residual blocks; (2) hashing layer for binarization; (3) supervised retrieval loss function based on neighbourhood component  ...  Most existing hashing methods first encode the images as a vector of hand-crafted features followed by a separate binarization step to generate hash codes.  ...  architecture for deep residual hashing (DRH) with a hash layer.  ... 
arXiv:1612.05400v1 fatcat:5gao7f3vqzcn5cvwv2jn4snzki

Attention-Aware joint Location Constraint Hashing for Multi-Label Image Retrieval

Yingqi Zhang, Yong Feng, Jiaxing Shang, Mingliang Zhou, Baohua Qiang
2019 IEEE Access  
Third, to learn more distinguishable hash codes, we leverage an attention sub-network to identify the approximate regions of the objects in an image so that the extracted features can mainly focus on the  ...  First, we leverage an adjacency matrix to record the relative location relationship among multiple objects.  ...  Deep Attention-guided Hashing (DAgH) [27] leveraged an attention network to obtain the salient regions of the images which are then used to guide the hash codes learning.  ... 
doi:10.1109/access.2019.2962084 fatcat:gwohyfwdrzg2hh4iv73sgk5jie

Scalable and Sustainable Deep Learning via Randomized Hashing [article]

Ryan Spring, Anshumali Shrivastava
2016 arXiv   pre-print
We present a novel hashing based technique to drastically reduce the amount of computation needed to train and test deep networks.  ...  Our new algorithm for deep learning reduces the overall computational cost of forward and back-propagation by operating on significantly fewer (sparse) nodes.  ...  A notable line of work [3, 27, 28] improved the accuracy of neural networks by only updating the neurons with the highest activations.  ... 
arXiv:1602.08194v2 fatcat:fo2pjpzsivgzxnmxurmlalfvda

Learning to Hash with Convolutional Network for Multi-label Remote Sensing Image Retrieval

Marwa Moustafa, National Authority for Remote Sensing and Space Sciences, Sayed Ahmed, Amal Hamed, National Authority for Remote Sensing and Space Sciences, National Authority for Remote Sensing and Space Sciences
2020 International Journal of Intelligent Engineering and Systems  
The hash network stacked two fully connected layers aimed to learn multiple hashing functions to encode the feature vector into a compact hash code.  ...  Specifically, the proposed MLRSIR-NET composed of two main sub-networks: multi-level feature extraction and deep hash.  ...  Conflicts of Interest The authors declare no conflict of interest. Author Contributions  ... 
doi:10.22266/ijies2020.1031.47 fatcat:56mnxvw2evhjhnovwrxvoaflum

Query-adaptive Image Retrieval by Deep Weighted Hashing [article]

Jian Zhang, Yuxin Peng
2017 arXiv   pre-print
Some deep hashing methods have achieved promising results by taking advantage of the strong representation power of deep networks recently.  ...  First, a novel deep hashing network is proposed to learn the hash codes and corresponding class-wise weights jointly, so that the learned weights can reflect the importance of different hash bits for different  ...  In the proposed deep network, a hash layer and a class-wise weight layer are designed, of which the hash layer generates hash codes, while the class-wise weight layer learns the classwise weights for different  ... 
arXiv:1612.02541v2 fatcat:zzvt74fgnzhjbcwjdkxoppeulm

Efficient Visual Recognition

Li Liu, Matti Pietikäinen, Jie Qin, Wanli Ouyang, Luc Van Gool
2020 International Journal of Computer Vision  
(SSN) which learns to select a single normalizer for each normalization layer of a deep network to improve interpretability and inference speed over SN.  ...  The paper "Anchor based Selfensembling for Semisupervised Deep Pairwise Hashing" by Xiaoshuang Shi, Zhenhua Guo, Fuyong Xing, Yun Liang, Lin Yang proposes a label efficient deep hashing method by leveraging  ... 
doi:10.1007/s11263-020-01351-w fatcat:mbcq6shmerbo5njayscgb3t4rq

One Loss for All: Deep Hashing with a Single Cosine Similarity based Learning Objective [article]

Jiun Tian Hoe and Kam Woh Ng and Tianyu Zhang and Chee Seng Chan and Yi-Zhe Song and Tao Xiang
2021 arXiv   pre-print
The result is an one-loss deep hashing model that removes all the hassles of tuning the weights of various losses.  ...  Importantly, extensive experiments show that our model is highly effective, outperforming the state-of-the-art multi-loss hashing models on three large-scale instance retrieval benchmarks, often by significant  ...  Acknowledgments and Disclosure of Funding  ... 
arXiv:2109.14449v1 fatcat:o4hz2nt22zbwjojbdamrcdlhuu

DREW: Efficient Winograd CNN Inference with Deep Reuse

Ruofan Wu, Feng Zhang, Jiawei Guan, Zhen Zheng, Xiaoyong Du, Xipeng Shen
2022 Proceedings of the ACM Web Conference 2022  
Convolutional neural networks (CNNs), which are deep learning representatives, are among the most popular neural networks in Web systems. However, CNN employs a high degree of computing.  ...  Fortunately, a minimal filtering algorithm called Winograd can reduce convolution calculations by minimizing multiplication operations.  ...  offline by leveraging the similarities among weights.  ... 
doi:10.1145/3485447.3511985 fatcat:qle2iycttrebjpcce4r5izehja

Feature Pyramid Hashing [article]

Yifan Yang, Libing Geng, Hanjiang Lai, Yan Pan, Jian Yin
2019 arXiv   pre-print
Inspired by the feature pyramids of convolutional neural network, a vertical pyramid is proposed to capture the high-layer features and a horizontal pyramid combines multiple low-layer features with structural  ...  In recent years, deep-networks-based hashing has become a leading approach for large-scale image retrieval.  ...  In recent years, inspired by the significant achievements of deep neural networks, learning the hash codes with deep neural networks (deep hashing) has become a novel stream of supervised hashing methods  ... 
arXiv:1904.02325v1 fatcat:wjp6syjdufd5pezneepvglifaa

Deep Momentum Uncertainty Hashing [article]

Chaoyou Fu, Guoli Wang, Xiang Wu, Qian Zhang, Ran He
2021 arXiv   pre-print
The discrepancy of each bit indicates the uncertainty of the hashing network to the approximate output of that bit.  ...  Specifically, we model bit-level uncertainty via measuring the discrepancy between the output of a hashing network and that of a momentum-updated network.  ...  Generally, the last layer of a neural network is leveraged to output binary hashing codes [12] .  ... 
arXiv:2009.08012v3 fatcat:2fej6v2a2zezjenwgtnhpl4eka

Deep Semantic Hashing with Generative Adversarial Networks

Zhaofan Qiu, Yingwei Pan, Ting Yao, Tao Mei
2017 Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval - SIGIR '17  
Specifically, a novel deep semantic hashing with GANs (DSH-GANs) is presented, which mainly consists of four components: a deep convolution neural networks (CNN) for learning image representations, an  ...  Our framework also achieves superior results when compared to state-of-the-art deep hash models.  ...  Semi-supervised hashing approaches attempt to improve the quality of hash codes by leveraging supervised information into learning procedure.  ... 
doi:10.1145/3077136.3080842 dblp:conf/sigir/QiuPYM17 fatcat:h5jc4hkgeng5rnhbzu6gou4f3e
« Previous Showing results 1 — 15 out of 6,123 results