Learning Compact Binary Descriptors with Unsupervised Deep Neural Networks

Lin, Kevin; Lu, Jiwen; Chen, Chu-Song; Zhou, Jie

doi:10.1109/cvpr.2016.133

Cited by 335 publications

(263 citation statements)

References 41 publications

Supporting

Mentioning

263

Contrasting

Order By: Relevance

“…We compare DDH with other hashing methods, such as LSH [1], ITQ [9], HS [31], Spectral hashing (SpeH) [45], Spherical hashing (SphH) [14], KMH [12], Deep Hashing (DH) [24] and DeepBit [23] , Semi-supervised PCAH [44] on the CIFAR-10 dataset and NUS-WIDE. We set the K 1 =15 and K 2 =6 to construct labels, and the learning rate as 0.001, λ 1 =15, λ 2 =0.00001 and batch-size=128.…”

Section: Results On Image Retrievalmentioning

confidence: 99%

“…To the best of our knowledge, DeepBit [23] is the first to propose a deep neural network to learn binary descriptors in an unsupervised manner, by enforcing three criteria on binary codes. It achieves the state-of-art performance for image retrieval, but DeepBit does not consider the data distribution in the original image space.…”

Section: Introductionmentioning

confidence: 99%

“…On the other hand, data-dependent hashing methods learn hash functions by exploring the distribution of the training data and therefore, they are also called learning to hashing methods (L2H) [43]. A lot of L2H methods have been proposed, such as Spectral hashing (SpeH) [45], iterative quantization (ITQ) [9], Multiple Feature Hashing (MFH) [36], Quantization-based Hashing (QBH) [32], K-means Hashing (KMH) [12], DH [24], DPSH [20], DeepBit [23], etc. Actually, those methods can be further divided into two categories: supervised methods and unsupervised methods.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep Discrete Hashing with Self-supervised Pairwise Labels

Song

Fan

et al. 2017

Machine Learning and Knowledge Discovery in Databases

View full text Add to dashboard Cite

Abstract. Hashing methods have been widely used for applications of large-scale image retrieval and classification. Non-deep hashing methods using handcrafted features have been significantly outperformed by deep hashing methods due to their better feature representation and end-to-end learning framework. However, the most striking successes in deep hashing have mostly involved discriminative models, which require labels. In this paper, we propose a novel unsupervised deep hashing method, named Deep Discrete Hashing (DDH), for large-scale image retrieval and classification. In the proposed framework, we address two main problems: 1) how to directly learn discrete binary codes? 2) how to equip the binary representation with the ability of accurate image retrieval and classification in an unsupervised way? We resolve these problems by introducing an intermediate variable and a loss function steering the learning process, which is based on the neighborhood structure in the original space. Experimental results on standard datasets (CIFAR-10, NUS-WIDE, and Oxford-17) demonstrate that our DDH significantly outperforms existing hashing methods by large margin in terms of mAP for image retrieval and object recognition. Code is available at https://github.com/htconquer/ddh.

show abstract

Section: Results On Image Retrievalmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Discrete Hashing with Self-supervised Pairwise Labels

Song

Fan

et al. 2017

Machine Learning and Knowledge Discovery in Databases

View full text Add to dashboard Cite

show abstract

“…However, most state-of-the-art unsupervised deep hashing algorithms [39]- [41] suffer from severe performance degradation due to lack of label information, e.g. DH [39] and Deepbit [40] use quantisation loss and evenly distribution loss, which can only make the binary codes have large variance, but can not get the real distribution of the data, UH-BDNN [41] adds a reconstruction layer after the binary code layer, which can obtain the semantics of the images, but still can not get the category information, which is more important than the semantics for image retrieval, and the experimental results reported in their original literatures and this paper can also show this point. Therefore, it is necessary to find out the visual feature distribution for unsupervised hashing.…”

Section: Introductionmentioning

confidence: 99%

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

“…For object matching, Lin et al [28] proposed an unsupervised learning to learn a compact binary descriptor by leveraging an iterative training scheme. More closely related to our work is the method of Zhou et al [53], which exploits cycleconsistency with a 3D CAD model [35] as a supervisory signal to train a deep network for semantic correspondence.…”

Section: Related Workmentioning

confidence: 99%

FCSS: Fully Convolutional Self-Similarity for Dense Semantic Correspondence

Kim

Min

Ham

et al. 2019

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

We present a descriptor, called fully convolutional selfsimilarity (FCSS), for dense semantic correspondence. To robustly match points among different instances within the same object class, we formulate FCSS using local selfsimilarity (LSS) within a fully convolutional network. In contrast to existing CNN-based descriptors, FCSS is inherently insensitive to intra-class appearance variations because of its LSS-based structure, while maintaining the precise localization ability of deep neural networks. The sampling patterns of local structure and the self-similarity measure are jointly learned within the proposed network in an end-to-end and multi-scale manner. As training data for semantic correspondence is rather limited, we propose to leverage object candidate priors provided in existing image datasets and also correspondence consistency between object pairs to enable weakly-supervised learning. Experiments demonstrate that FCSS outperforms conventional handcrafted descriptors and CNN-based descriptors on various benchmarks.

show abstract

Learning Compact Binary Descriptors with Unsupervised Deep Neural Networks

Cited by 335 publications

References 41 publications

Deep Discrete Hashing with Self-supervised Pairwise Labels

Deep Discrete Hashing with Self-supervised Pairwise Labels

Unsupervised Deep Hashing With Pseudo Labels for Scalable Image Retrieval

FCSS: Fully Convolutional Self-Similarity for Dense Semantic Correspondence

Contact Info

Product

Resources

About