Supervised Learning of Semantics-Preserving Hash via Deep Convolutional Neural Networks

Yang, Huei-Fang; Lin, Kevin; Chen, Chu-Song

doi:10.1109/tpami.2017.2666812

Cited by 283 publications

(201 citation statements)

References 46 publications

Supporting

Mentioning

201

Contrasting

Order By: Relevance

“…Through optimising the sum of reconstruction loss, independence constraint loss, and balance constraint loss, BDNN can get a little improvement with respect to DH. The recent state-of-art method is Supervised Semantic-preserving Deep Hashing (SSDH) [26], which assumes that the semantic labels are governed by several latent attributes. Based on this assumption, SSDH constructs hash function as a latent layer in a deep network and learns the binary codes by minimising an objective function combined by classification error, independent property, and balance property like those in method BDNN.…”

Section: A Deep Hashing Methodsmentioning

confidence: 99%

“…To solve the problem of NP-hard optimisation and computational complexity in graph based method, Asymmetric Discrete Graph Hashing (ADGH) [49] was proposed by preserving the asymmetric discrete constraint and building an asymmetric affinity matrix to learn compact binary codes. Additionally, SSDH [26] constructs hash functions as a latent layer in a deep network and the binary codes are learned by minimising an objective function defined over classification error.…”

Section: Related Workmentioning

confidence: 99%

“…Canonical Correlation Analysis ITQ (CCA-ITQ) [22], Supervised Semantics-preserving Deep Hashing (SSDH) [26], Quantization Based Hashing (QBH) [27], which are all trained with labelled samples and can achieve great improvement in performance. However, these works are all supervised with real labels.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Unsupervised Deep Hashing With Pseudo Labels for Scalable Image Retrieval

Zhang

Liu

Long

et al. 2018

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Abstract-In order to achieve efficient similarity searching, hash functions are designed to encode images into lowdimensional binary codes with the constraint that similar features will have a short distance in the projected Hamming space. Recently, deep learning-based methods have become more popular, and outperform traditional non-deep methods. However, without label information, most state-of-the-art unsupervised deep hashing (DH) algorithms suffer from severe performance degradation for unsupervised scenarios. One of the main reasons is that the ad-hoc encoding process cannot properly capture the visual feature distribution. In this paper, we propose a novel unsupervised framework that has two main contributions: 1) we convert the unsupervised DH model into supervised by discovering pseudo labels; 2) the framework unifies likelihood maximization, mutual information maximization, and quantization error minimization so that the pseudo labels can maximumly preserve the distribution of visual features. Extensive experiments on three popular data sets demonstrate the advantages of the proposed method, which leads to significant performance improvement over the state-of-the-art unsupervised hashing algorithms. Index Terms-Image retrieval, unsupervised hashing, pseudo labels.

show abstract

Section: A Deep Hashing Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Unsupervised Deep Hashing With Pseudo Labels for Scalable Image Retrieval

Zhang

Liu

Long

et al. 2018

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

show abstract

“…In order to push the final real activations towards the extremities of the sigmoid range, and inspired by [9], we use a second loss whose goal is to maximize the sum of the squared errors between the output-layer activations and the value 0.5:…”

Section: Metric-learning Based Deep Hashing Networkmentioning

confidence: 99%

Metric-Learning-Based Deep Hashing Network for Content-Based Retrieval of Remote Sensing Images

Roy

Sangineto

Demir

et al. 2021

IEEE Geosci. Remote Sensing Lett.

View full text Add to dashboard Cite

Hashing methods have recently been shown to be very effective in retrieval of Remote Sensing (RS) images due to their computational efficiency and fast search speed. Common hashing methods in RS are based on hand-crafted features on top of which they learn a hash function which provides the final binary codes. However, these features are not optimized for the final task (i.e., retrieval using binary codes). On the other hand, modern Deep Neural Networks (DNNs) have shown an impressive success in learning optimized features for a specific task in an end-to-end fashion. Unfortunately, typical RS datasets are composed of only a small number of labeled samples, which makes the training (or fine-tuning) of big DNNs problematic and prone to overfitting. To address this problem, in this paper we introduce a metric-learning based hashing network, which: 1) Implicitly uses a big, pre-trained DNN as an intermediate representation step without the need of re-training or fine-tuning; 2) Learns a semantic-based metric space where the features are optimized for the target retrieval task; 3) Computes compact binary hash codes for fast search. Experiments carried out on two RS benchmarks highlight that the proposed network significantly improves the retrieval performance under the same retrieval time when compared to the state-of-the-art hashing methods in RS.Index Terms-deep hashing, metric learning, content based image retrieval, remote sensing.

show abstract

“…Supervised Discrete Hashing (SDH) [18] improves retrieval accuracy by integrating classify and binary codes generation during training. Recently, many deep hashing methods are proposed [3,6,10,29,[32][33][34]. According to the forms of similarity preserving manners, two supervised information are widely used: 1) the pairwise-based methods and 2) the triplet-based methods.…”

Section: Related Workmentioning

confidence: 99%

Deep Policy Hashing Network with Listwise Supervision

Wang

Lai

Yang

et al. 2019

Proceedings of the 2019 on International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

Deep-networks-based hashing has become a leading approach for large-scale image retrieval, which learns a similarity-preserving network to map similar images to nearby hash codes. The pairwise and triplet losses are two widely used similarity preserving manners for deep hashing. These manners ignore the fact that hashing is a prediction task on the list of binary codes. However, learning deep hashing with listwise supervision is challenging in 1) how to obtain the rank list of whole training set when the batch size of the deep network is always small and 2) how to utilize the listwise supervision. In this paper, we present a novel deep policy hashing architecture with two systems are learned in parallel: a query network and a shared and slowly changing database network.The following three steps are repeated until convergence: 1) the database network encodes all training samples into binary codes to obtain whole rank list, 2) the query network is trained based on policy learning to maximize a reward that indicates the performance of the whole ranking list of binary codes, e.g., mean average precision (MAP), and 3) the database network is updated as the query network. Extensive evaluations on several benchmark datasets show that the proposed method brings substantial improvements over state-of-the-art hashing methods.

show abstract

Supervised Learning of Semantics-Preserving Hash via Deep Convolutional Neural Networks

Cited by 283 publications

References 46 publications

Unsupervised Deep Hashing With Pseudo Labels for Scalable Image Retrieval

Unsupervised Deep Hashing With Pseudo Labels for Scalable Image Retrieval

Metric-Learning-Based Deep Hashing Network for Content-Based Retrieval of Remote Sensing Images

Deep Policy Hashing Network with Listwise Supervision

Contact Info

Product

Resources

About