Learning-based Hashing Research Articles

Deep Attention Fusion Hashing (DAFH) Model for Medical Image Retrieval.

To address this, we propose a novel deep learning-based hashing model, the Deep Attention Fusion Hashing (DAFH) model, which integrates advanced attention mechanisms with medical imaging data. The DAFH model enhances retrieval performance by integrating multi-modality medical imaging data and employing attention mechanisms to optimize the feature extraction process. Utilizing multimodal medical image data from the Cancer Imaging Archive (TCIA), this study constructed and trained a deep hashing network that achieves high-precision classification of various cancer types. At hash code lengths of 16, 32, and 48 bits, the model respectively attained Mean Average Precision (MAP@10) values of 0.711, 0.754, and 0.762, highlighting the potential and advantage of the DAFH model in medical image retrieval. The DAFH model demonstrates significant improvements in the efficiency and accuracy of medical image retrieval, proving to be a valuable tool in clinical settings.

Efficient indexing method based on autoencoder for large-scale data products

Abstract Traditional indexing methods fail to meet the power industry’s needs for rapid content retrieval and collaborative management of heterogeneous data. This study proposes a learning-based hashing function using autoencoders, which compresses raw data into binary hash codes by learning data features and reconstruction. Extensive experiments demonstrate its advantages in reducing collision rates and computation time. Compared to various models, this method shows clear benefits in search and index creation times, better addressing the indexing requirements of power data and supporting intelligent processing in power systems.

MLMQ-IR: Multi-label multi-query image retrieval based on the variance of Hamming distance

Image retrieval (IR) methods extract the most relevant images to the query images from an image database. The existing IR methods, which retrieve images with a low degree of similarity, use computationally expensive approaches. In this study, a novel Multi-Label Multi-Query IR (MLMQ-IR) method based on the variance of Hamming distance is presented for the query of multiple images having multiple labels. The MLMQ-IR uses deep learning-based hashing code generation with ResNet50 structure. The variance evaluates the minimum variation of the distances between the query and database images, providing the trade-off images according to the center of Pareto space. Moreover, the MLMQ-IR exploits a new Triple Loss Multi-Label Hashing (TLMH) depending on binary cross-entropy loss and bit-balance loss functions. The MLMQ-IR is compared with recent multi-label and multi-query methods through MIRFLICKR-25 K, MS-COCO and NUS-WIDE datasets in terms of three well-known metrics, and the methods are ranked with succussed sorting. As a result, the MLMQ-IR method has the best avg. mean rank with 1.86. The results manifest that the MLMQ-IR provides the most similar retrieved images to the query images owing to utilizing the variance which is the efficient and fast IR approach.

BH2I-GAN: Bidirectional Hash_code-to-Image Translation using Multi-Generative Multi-Adversarial Nets

Given the benefits of high retrieval efficiency and low storage cost, hashing method has received an increasing attention. In particular, deep learning-based hashing has been widely used in data mining and information retrieval. However, almost all the existing methods only achieve the goal of high retrieval precision, and limit the evaluation of hashing methods to objective aspect. In this paper, we propose a novel bidirectional hash_code-to-image translation model by using multi-generative multi-adversarial nets to reduce storage cost truly and obtain satisfactory user acceptance on the basis of high retrieval precision. Firstly, we propose supervised manifold metric to reduce Hamming distance between similar instances while increasing the Hamming distance between dissimilar instances, which have been proved to be helpful for high retrieval precision and good user acceptance. Then, we utilize multi-generative and multi-adversarial networks to construct hash mapping and inverse hash generation. During inverse generation, theoretical analysis is conducted to show that inverse hash network can avoid unstable training and mode collapse. Besides, we prove that Poisson distribution induced by hash codes can be initialized as generative distribution to fit real distribution. Experimental results show that our method outperforms several state-of-the-art approaches on three popular datasets. Specifically, ours yields average about 9.3% increment in Mean Average Precision(MAP) on three datasets, and achieves over 90% user satisfaction. Besides, it successfully reduces storage cost by 1,634 times in COCO 2017 large-scale dataset.

Learning to hash: a comprehensive survey of deep learning-based hashing methods

Learning to hash: a comprehensive survey of deep learning-based hashing methods

Ranking-Based Deep Hashing Network for Image Retrieval

In large-scale image retrieval, the deep learning-based hashing methods have significantly progressed. However, most of the existing deep hashing methods still have the problems of low feature learning efficiency and weak ranking relationship discrimination. To remedy these problems, a novel Ranking-based Deep Hashing Network (RDHN) is proposed for image retrieval in this paper, which integrates the feature learning module and hash learning module into a unified deep hashing network framework to jointly learn a powerful hash function so that the raw images can be mapped to discrete hash codes with significant discrimination. Specifically, a novel difference convolution is designed based on edge detection operators, and then it is uniquely applied to the first convolutional layer of the convolutional neural network (CNN), which can take advantage of the sensitive characteristics of edge detection operators for edge information to extract richer image edge information. Meanwhile, in hash learning, the ranking metric Mean Average Precision (MAP) is optimized using the idea of scaling, and then a ranking loss function based on MAP is carefully designed to enhance the neighborhood ranking capabilities of the hash codes. Furthermore, to reduce the quantification error, a quantization loss function is also designed. Finally, the ranking loss function is combined with the quantization loss function to form the objective function. The proposed method can generate high-quality discrete hash codes while learning to preserve ranking information, effectively improving retrieval performance. Extensive experimental results on three widely used benchmark datasets demonstrate that the proposed method outperforms the state-of-the-art hashing approaches.

Dual Attention Triplet Hashing Network for Image Retrieval.

In recent years, learning-based hashing techniques have proven to be efficient for large-scale image retrieval. However, since most of the hash codes learned by deep hashing methods contain repetitive and correlated information, there are some limitations. In this paper, we propose a Dual Attention Triplet Hashing Network (DATH). DATH is implemented with two-stream ConvNet architecture. Specifically, the first neural network focuses on the spatial semantic relevance, and the second neural network focuses on the channel semantic correlation. These two neural networks are incorporated to create an end-to-end trainable framework. At the same time, in order to make better use of label information, DATH combines triplet likelihood loss and classification loss to optimize the network. Experimental results show that DATH has achieved the state-of-the-art performance on benchmark datasets.

A supervised deep convolutional based bidirectional long short term memory video hashing for large scale video retrieval applications

Recently, large scale video content retrieval has gained more attention due to a large amount of user-generated images and video content available over the internet. Hashing is one of the effective techniques that encode the high dimensional features vectors into compact binary codes. The aim of hashing is to generate the short binary codes and map the similar hash code values to retrieve similar video from the database with minimum distance measure. Deep learning-based hashing networks are employed to learn the representative video feature to estimate the hash functions. However, the existing hashing approaches fail due to the frame-level feature representation and does not well exploits the effective temporal features in visual search. Furthermore, significant loss of features during the dimensionality reduction step causes low accuracy. Hence, it is essential to develop a deep learning-based hashing framework that should exploits both the strong spatial and temporal features for the scalable video search. The main objective is to learn the high dimensional features from the entire video and derive compact binary codes to retrieve the similar videos for the input query sequence. In this paper, we propose a joint network model of supervised stacked heterogeneous convolutional multi-kernel (Stacked HetConv-MK)-bidirectional Long Short Term Memory (BiDLSTM) network model that effectively encodes the rich structural as well as the discriminative features from the video sequence to estimate the compact binary codes. Initially, the video frames are passed to the stacked convolution networks with heterogeneous convolutional kernel size and residual learning to extract the spatial features at different views from the video sequences and to improve the learning efficiency. Then, the bidirectional network computes the sequence in both forward and backward directions and obtains the series of hidden state output. Finally, the fully connected structure with an activation unit performs hashing to learn the multiple codes for each video. Experimental analysis is performed on three datasets and the result shows a better accuracy measure than the other state-of-art approaches.

Quantization-based hashing with optimal bits for efficient recommendation

Recommendation technique has been widely applied in e-commerce systems, but the efficiency becomes challenging due to the growing scale of users and items. In recent years, several hashing-based recommendation frameworks were proposed to solve the efficiency issue successfully by representing users and items with binary codes. These hashing methods consist of two types: two-stage hashing and learning-based hashing. In this paper, we focus on putting forward to a two-stage hashing called quantization-based hashing (QBH) to alleviate the efficiency bottleneck and improve the recommendation accuracy as well. To be specific, we propose the QBH that consists of similarity quantization and norm quantization. To improve the accuracy performance, we search the optimal bits of quantization by minimizing a quantization loss function. We finally evaluate the proposed method on three public datasets to show its superiority on recommendation accuracy over other two-stage hashing methods and advantage on recommendation efficiency over the state-of-the-art recommender systems.

Local Feature Hashing With Binary Auto-Encoder for Face Recognition

The learning-based hashing has recently made encouraging progress in face recognition. However, most existing hashing methods disregard the discrete constraint during optimization, inducing the accumulated quantization errors. In this work, we develop an effective learning-based hashing model, namely local feature hashing with binary auto-encoder (LFH-BAE), to directly learn local binary descriptors in the Hamming space. It attempts to exploit structure factors to well reconstruct the face image from binary codes. Specifically, we first introduce a binary auto-encoder to learn a hashing function to project each face region into high-quality binary codes. Since the original problem is a tricky combinational function, we then present a softened version to decompose it into separate tractable sub-problems. Next, we propose an effective alternating algorithm based on the augmented Lagrange method (ALM) to solve these sub-problems, which helps to generate strong discriminative and excellent robust binary codes. Moreover, we utilize the discrete cyclic coordinate descent (DCC) method to optimize binary codes to reduce the loss of useful information. Lastly, we cluster and pool the obtained binary codes, and construct a histogram feature as the final face representation for each image. Extensive experimental results on four public datasets including FERET, CAS-PEAL-R1, LFW and PaSC show that our LFH-BAE is superior to most state-of-the-art face recognition algorithms.

Deep Progressive Hashing for Image Retrieval

Hashing is a widely adopted method based on an approximate nearest neighbor search and is used in large-scale image retrieval tasks. Conventional learning-based hashing algorithms employ end-to-end representation learning, which is a one-off technique. Because of the tradeoff between efficiency and performance, conventional learning-based hashing methods must sacrifice code length to improve performance, which increases their computational complexity. To improve the efficiency of binary codes, motivated by the “nonsalient-to-salient” attention scheme of humans, we propose a recursive hashing mechanism that maps progressively expanded salient regions to a series of binary codes. These salient regions are generated by a conventional saliency model based on bottom-up saliency-driven attention and a semantic-guided saliency model based on top-down task-driven attention. After obtaining a series of salient regions, we perform long-range temporal modeling of salient regions using a graph-based recurrent deep network to obtain more refined representative features. The later output nodes inherit aggregated information from all previous nodes and extract discriminative features from more salient regions. Therefore, this network possesses more significant information and satisfactory scalability. The proposed recursive hashing neural network, optimized by a triplet ranking loss, is end-to-end trainable. Extensive experimental results from several image retrieval benchmarks show the scalability of our method and demonstrate its strong performance compared with state-of-the-art methods.

Learning rotation-invariant binary codes for efficient object detection from remote sensing images

Object detection is one of the most important research topics for remote sensing image (RSI) analysis. Because of the rapid development of satellite imaging technology, the resolution of RSI has increased dramatically, which brings a great challenge to the efficiency of RSI-based object detection. We present an object detection model, which greatly improves the detection speed by using learning-based hashing techniques. Specifically, first, a selective search method is used to generate many high quality object proposals that may contain objects. Then for each proposal, we learn rotation-invariant binary codes, which can handle the objects with arbitrary orientations in RSI, to quickly eliminate most nonobject proposals in Hamming space. And finally, the object detection task can be achieved by classifying the left (very limited amount of) proposals with more discriminating classification model. Experimental evaluations on a public very high-resolution remote sensing dataset show the superiority of the proposed method. Specifically, the rotation-invariant learning metric has been combined with three popular hashing methods, and performance improvements have been obtained for all the cases. Furthermore, comparisons with deep learning-based detectors show that hash-based methods fail to achieve state-of-the-art detection performance, but when the graphics processing unit is inaccessible, it achieves a good compromise between performance and runtime.

Unsupervised Semantic-Preserving Adversarial Hashing for Image Search.

Hashing plays a pivotal role in nearest-neighbor searching for large-scale image retrieval. Recently, deep learning-based hashing methods have achieved promising performance. However, most of these deep methods involve discriminative models, which require large-scale, labeled training datasets, thus hindering their real-world applications. In this paper, we propose a novel strategy to exploit the semantic similarity of the training data and design an efficient generative adversarial framework to learn binary hash codes in an unsupervised manner. Specifically, our model consists of three different neural networks: an encoder network to learn hash codes from images, a generative network to generate images from hash codes, and a discriminative network to distinguish between pairs of hash codes and images. By adversarially training these networks, we successfully learn mutually coherent encoder and generative networks, and can output efficient hash codes from the encoder network. We also propose a novel strategy, which utilizes both feature and neighbor similarities, to construct a semantic similarity matrix, then use this matrix to guide the hash code learning process. Integrating the supervision of this semantic similarity matrix into the adversarial learning framework can efficiently preserve the semantic information of training data in Hamming space. The experimental results on three widely used benchmarks show that our method not only significantly outperforms several state-of-the-art unsupervised hashing methods, but also achieves comparable performance with popular supervised hashing methods.

Angular Deep Supervised Hashing for Image Retrieval

Deep learning based image hashing methods learn hash codes by using powerful feature extractors and nonlinear transformations to achieve highly efficient image retrieval. For most end-to-end deep hashing methods, the supervised learning process relies on pair-wise or triplet-wise information to provide an internal relationship of similarity data. However, the use of pair-wise and triplet loss function is limited not only by expensive training costs but also by quantization errors. In this paper, we propose a novel semantic learning based hashing method for image retrieval to optimize the deep features structure in the hash space from a perspective of angular view. Specifically, we proposed an angular hashing loss function that explicitly improve intra-class compactness and inter-class separability between features in hash space. Geometrically, angular hashing loss can be regarded as imposing hash constraints on hypersphere manifold. In order to solve the training problem on the multi-label case, we further designed a dynamic Softmax training strategy that can directly train the network using gradient descent method. Extensive experiments on two well-known datasets of CIFAR-10 and NUS-WIDE demonstrate that the proposed Angular Deep Supervised Hashing (ADSH) method can generate high-quality and compact binary codes, which can achieve state-of-the-art performance as compared with conventional image hashing and deep learning-based hashing methods.

Deep Attention-Guided Hashing

With the rapid growth of multimedia data (e.g., image, audio and video etc.) on the web, learning-based hashing techniques such as Deep Supervised Hashing (DSH) have proven to be very efficient for large-scale multimedia search. The recent successes seen in Learning-based hashing methods are largely due to the success of deep learning-based hashing methods. However, there are some limitations to previous learning-based hashing methods (e.g., the learned hash codes containing repetitive and highly correlated information). In this paper, we propose a novel learning-based hashing method, named Deep Attention-guided Hashing (DAgH). DAgH is implemented using two stream frameworks. The core idea is to use guided hash codes which are generated by the hashing network of the first stream framework (called first hashing network) to guide the training of the hashing network of the second stream framework (called second hashing network). Specifically, in the first network, it leverages an attention network and hashing network to generate the attention-guided hash codes from the original images. The loss function we propose contains two components: the semantic loss and the attention loss. The attention loss is used to punish the attention network to obtain the salient region from pairs of images; in the second network, these attention-guided hash codes are used to guide the training of the second hashing network (i.e., these codes are treated as supervised labels to train the second network). By doing this, DAgH can make full use of the most critical information contained in images to guide the second hashing network in order to learn efficient hash codes in a true end-to-end fashion. Results from our experiments demonstrate that DAgH can generate high quality hash codes and it outperforms current state-of-the-art methods on three benchmark datasets, CIFAR-10, NUS-WIDE, and ImageNet.

Object-Location-Aware Hashing for Multi-Label Image Retrieval via Automatic Mask Learning.

Learning-based hashing is a leading approach of approximate nearest neighbor search for large-scale image retrieval. In this paper, we develop a deep supervised hashing method for multi-label image retrieval, in which we propose to learn a binary "mask" map that can identify the approximate locations of objects in an image, so that we use this binary "mask" map to obtain length-limited hash codes which mainly focus on an image's objects but ignore the background. The proposed deep architecture consists of four parts: 1) a convolutional sub-network to generate effective image features; 2) a binary "mask" sub-network to identify image objects' approximate locations; 3) a weighted average pooling operation based on the binary "mask" to obtain feature representations and hash codes that pay most attention to foreground objects but ignore the background; and 4) the combination of a triplet ranking loss designed to preserve relative similarities among images and a cross entropy loss defined on image labels. We conduct comprehensive evaluations on four multi-label image data sets. The results indicate that the proposed hashing method achieves superior performance gains over the state-of-the-art supervised or unsupervised hashing baselines.

Error Correcting Input and Output Hashing.

Most learning-based hashing algorithms leverage sample-to-sample similarities, such as neighborhood structure, to generate binary codes, which achieve promising results for image retrieval. This type of methods are referred to as instance-level encoding. However, it is nontrivial to define a scalar to represent sample-to-sample similarity encoding the semantic labels and the data structure. To address this issue, in this paper, we seek to use a class-level encoding method, which encodes the class-to-class relationship, to take the semantic information of classes into consideration. Based on these two encodings, we propose a novel framework, error correcting input and output (EC-IO) coding, which does class-level and instance-level encoding under a unified mapping space. Our proposed model contains two major components, which are distribution preservation and error correction. With these two components, our model maps the input feature of samples and the output code of classes into a unified space to encode the intrinsic structure of data and semantic information of classes simultaneously. Under this framework, we present our hashing model, EC-IO hashing (EC-IOH), by approximating the mapping space with the Hamming space. Extensive experiments are conducted to evaluate the retrieval performance, and EC-IOH exhibits superior and competitive performances comparing with popular supervised and unsupervised hashing methods.

Learning Label Preserving Binary Codes for Multimedia Retrieval

Learning-based hashing has been researched extensively in the past few years due to its great potential in fast and accurate similarity search among huge volumes of multimedia data. In this article, we present a novel multimedia hashing framework, called Label Preserving Multimedia Hashing (LPMH) for multimedia similarity search. In LPMH, a general optimization method is used to learn the joint binary codes of multiple media types by explicitly preserving semantic label information. Compared with existing hashing methods which are typically developed under and thus restricted to some specific objective functions, the proposed optimization strategy is not tied to any specific loss function and can easily incorporate bit balance constraints to produce well-balanced binary codes. Specifically, our formulation leads to a set of Binary Integer Programming (BIP) problems that have exact solutions both with and without bit balance constraints. These problems can be solved extremely fast and the solution can easily scale up to large-scale datasets. In the hash function learning stage, the boosted decision trees algorithm is utilized to learn multiple media-specific hash functions that can map heterogeneous data sources into a homogeneous Hamming space for cross-media retrieval. We have comprehensively evaluated the proposed method using a range of large-scale datasets in both single-media and cross-media retrieval tasks. The experimental results demonstrate that LPMH is competitive with state-of-the-art methods in both speed and accuracy.

Financial credit risk assessment via learning-based hashing

Financial credit risk assessment via learning-based hashing

Quantization-based hashing: a general framework for scalable image and video retrieval

Nowadays, due to the exponential growth of user generated images and videos, there is an increasing interest in learning-based hashing methods. In computer vision, the hash functions are learned in such a way that the hash codes can preserve essential properties of the original space (or label information). Then the Hamming distance of the hash codes can approximate the data similarity. On the other hand, vector quantization methods quantize the data into different clusters based on the criteria of minimal quantization error, and then perform the search using look-up tables. While hashing methods using Hamming distance can achieve faster search speed, their accuracy is often outperformed by quantization methods with the same code length, due to the low quantization error and more flexible distance lookups. To improve the effectiveness of the hashing methods, in this work, we propose Quantization-based Hashing (QBH), a general framework which incorporates the advantages of quantization error reduction methods into conventional property preserving hashing methods. The learned hash codes simultaneously preserve the properties in the original space and reduce the quantization error, and thus can achieve better performance. Furthermore, the hash functions and a quantizer can be jointly learned and iteratively updated in a unified framework, which can be readily used to generate hash codes or quantize new data points. Importantly, QBH is a generic framework that can be integrated to different property preserving hashing methods and quantization strategies, and we apply QBH to both unsupervised and supervised hashing models as showcases in this paper. Experimental results on three large-scale unlabeled datasets (i.e., SIFT1M, GIST1M, and SIFT1B), three labeled datastes (i.e., ESPGAME, IAPRTC and MIRFLICKR) and one video dataset (UQ_VIDEO) demonstrate the superior performance of our QBH over existing unsupervised and supervised hashing methods.