538 Hits in 1.9 sec

Unsupervised similarity learning through Cartesian product of ranking references

Lucas Pascotti Valem, Daniel Carlos Guimarães Pedronette, Jurandy Almeida
2017 Pattern Recognition Letters  
In this scenario, similarity learning approaches capable of improving the effectiveness of retrieval in an unsupervised way are indispensable.  ...  A novel method, called Cartesian Product of Ranking References (CPRR), is proposed with this objective in this paper.  ...  In this paper, we present a novel unsupervised similarity learning method for improving the effectiveness of multimedia retrieval tasks, named as Cartesian product of ranking references (CPRR).  ... 
doi:10.1016/j.patrec.2017.10.013 fatcat:q2gbx7ofgvatrjyxriwtfz7t4y

SUBIC: A supervised, structured binary code for image search [article]

Himalaya Jain, Joaquin Zepeda, Patrick Pérez, Rémi Gribonval
2017 arXiv   pre-print
For large-scale visual search, highly compressed yet meaningful representations of images are essential.  ...  Yet, unlike binary hashing schemes, these unsupervised methods have not yet benefited from the supervision, end-to-end learning and novel architectures ushered in by the deep learning revolution.  ...  These two tasks consist of ranking database images according to their similarity to a given query image, where similarity is defined by membership in a given semantic category (category retrieval) or by  ... 
arXiv:1708.02932v1 fatcat:hzaxdpj3r5cdlpx5aoc3j26b34

Place recognition survey: An update on deep learning approaches [article]

Tiago Barros, Ricardo Pereira, Luís Garrote, Cristiano Premebida, Urbano J. Nunes
2021 arXiv   pre-print
Some lessons learned from this survey include: the importance of NetVLAD for supervised end-to-end learning; the advantages of unsupervised approaches in place recognition, namely for cross-domain applications  ...  As part of the localization system, place recognition has benefited from recent developments in other perception tasks such as place categorization or object recognition, namely with the emergence of deep  ...  The latent space, which is jointly learned on the two tasks, is used as a descriptor for segment retrieval.  ... 
arXiv:2106.10458v2 fatcat:hbw47qq2mjhsjhfb5t5vw4wfce

Improved Deep Hashing with Soft Pairwise Similarity for Multi-label Image Retrieval [article]

Zheng Zhang, Qin Zou, Yuewei Lin, Long Chen, Song Wang
2019 arXiv   pre-print
However, such similarity definition cannot reflect the similarity ranking for pairwise images that hold multiple labels.  ...  Hash coding has been widely used in the approximate nearest neighbor search for large-scale image retrieval.  ...  For multilabel retrieval, DSRH [25] tries to learn hash function by utilizing the ranking information of multi-level similarity, and proposes a surrogate losses to solve the optimization problem of ranking  ... 
arXiv:1803.02987v3 fatcat:koevkgmdtjgxpomsqwcaygi2zy

A Survey on Learning to Hash [article]

Jingdong Wang, Ting Zhang, Jingkuan Song, Nicu Sebe, Heng Tao Shen
2017 arXiv   pre-print
In this paper, we present a comprehensive survey of the learning to hash algorithms, categorize them according to the manners of preserving the similarities into: pairwise similarity preserving, multiwise  ...  Learning to hash is one of the major solutions to this problem and has been widely studied recently.  ...  ACKNOWLEDGEMENTS This work was partially supported by the National Nature Science Foundation of China No. 61632007.  ... 
arXiv:1606.00185v2 fatcat:j5mnu7lfmvby5pfkg5pffk2nae

Product Quantization Network for Fast Visual Search

Tan Yu, Jingjing Meng, Chen Fang, Hailin Jin, Junsong Yuan
2020 International Journal of Computer Vision  
Product quantization has been widely used in fast image retrieval due to its effectiveness of coding high-dimensional visual features.  ...  Meanwhile, by extending the triplet loss to the asymmetric triplet loss, we directly optimize the retrieval accuracy of the learned representation based on asymmetric similarity measurement.  ...  In the retrieval phase, we utilize the asymmetric similarity scores to rank the reference images in the dataset. Let denote by v q the output of embedding layer when the input is query image q.  ... 
doi:10.1007/s11263-020-01326-x fatcat:pjcb4wuzpff3pnggsvqeqi2hv4

Opponent Color And Curvelet Transform Based Image Retrieval System Using Genetic Algorithm

Yesubai Rubavathi Charles, Ravi Ramraj
2015 Zenodo  
Thus, the experimental results confirm that the proposed content based image retrieval system architecture attains better solution for image retrieval.  ...  The recent scenario in the issues of image retrieval is to reduce the semantic gap between user's preference and low level features.  ...  Unsupervised learning such as clustering, it is based on how the low level features are organized or clustered to retrieve the similar images.  ... 
doi:10.5281/zenodo.1110855 fatcat:brbdlrlpwvbftenhvblwsjqery

Deep Supervised Quantization by Self-Organizing Map

Min Wang, Wengang Zhou, Qi Tian, Junfu Pu, Houqiang Li
2017 Proceedings of the 2017 ACM on Multimedia Conference - MM '17  
With the supervised quantization loss, we minimize the differences on the maps between similar image pairs, and maximize the differences on the maps between dissimilar image pairs.  ...  The experiments on several public standard datasets prove the superiority of our approach over the existing ANN search methods.  ...  It maximizes the margin between the similar pairs and the dissimilar pairs, which intuitively guarantees learned binary codes to preserve the ranking orders of images.  ... 
doi:10.1145/3123266.3123415 dblp:conf/mm/WangZTPL17 fatcat:4kjtvt7e7bb5ravxckg6jpphgq

Unsupervised Neural Quantization for Compressed-Domain Similarity Search [article]

Stanislav Morozov, Artem Babenko
2019 arXiv   pre-print
We tackle the problem of unsupervised visual descriptors compression, which is a key ingredient of large-scale image retrieval systems.  ...  In more detail, we introduce a DNN architecture for the unsupervised compressed-domain retrieval, based on multi-codebook quantization.  ...  While originally appeared for the image retrieval problem, the quantization methods are extensively used to increase the efficiency in a wide range of tasks, e.g.  ... 
arXiv:1908.03883v1 fatcat:3aklfr6eiba5jndscoqwd6khjy

Statistical quantization for similarity search

Qi Wang, Guokang Zhu, Yuan Yuan
2014 Computer Vision and Image Understanding  
However, a common problem shared by the traditional quantizers is that during the out-of-sample extension process, the naive strategy considers only the similarities in Euclidean space without taking into  ...  Approximate nearest neighbor search has attracted much attention recently, which allows for fast query with a predictable sacrifice in search quality.  ...  For instance, in [28] the post-combination strategy is employed on the linear hash functions of different types of features for content-based image retrieval.  ... 
doi:10.1016/j.cviu.2014.03.002 fatcat:zzfvn54dcfai5nfzh4vb5eleru

Bloom Filters and Compact Hash Codes for Efficient and Distributed Image Retrieval

Andrea Salvi, Simone Ercoli, Marco Bertini, Alberto Del Bimbo
2016 2016 IEEE International Symposium on Multimedia (ISM)  
This paper presents a novel method for efficient image retrieval, based on a simple and effective hashing of CNN features and the use of an indexing structure based on Bloom filters.  ...  Experimental validation has been performed on three standard image retrieval datasets, outperforming state-of-the-art hashing methods in terms of precision, while the proposed indexing method obtains a  ...  Government is authorized to reproduce and distribute reprints for Governmental purpose notwithstanding any copyright annotation thereon.  ... 
doi:10.1109/ism.2016.0113 dblp:conf/ism/SalviEBB16 fatcat:fy7xmkbdyzderbmwjngzj6mm2q

Recent Advance in Content-based Image Retrieval: A Literature Survey [article]

Wengang Zhou, Houqiang Li, Qi Tian
2017 arXiv   pre-print
With the ignorance of visual content as a ranking clue, methods with text search techniques for visual retrieval may suffer inconsistency between the text words and visual content.  ...  Numerous techniques have been developed for content-based image retrieval in the last decade.  ...  It decomposes the feature space into a Cartesian product of low-dimensional subspaces and quantizes each sub-space individually.  ... 
arXiv:1706.06064v2 fatcat:m52xwsw5pzfzdbxo5o6dye2gde

Learning in High-Dimensional Multimedia Data: The State of the Art [article]

Lianli Gao, Jingkuan Song, Xingyi Liu, Junming Shao, Jiajun Liu, Jie Shao
2017 arXiv   pre-print
Machine learning is largely involved in multimedia applications of building models for classification and regression tasks etc., and the learning principle consists in designing the models based on the  ...  During the last decade, the deluge of multimedia data has impacted a wide range of research areas, including multimedia retrieval, 3D tracking, database management, data mining, machine learning, social  ...  While unsupervised hashing shows promise in retrieving neighbors based on metric distances, e.g., ℓ 2 distance, a variety of practical applications prefer semantically similar neighbors [86] .  ... 
arXiv:1707.02683v1 fatcat:wx5kr2hbybhlvbaz66646zjob4

Deep Multimodal Learning for Affective Analysis and Retrieval

Lei Pang, Shiai Zhu, Chong-Wah Ngo
2015 IEEE transactions on multimedia  
Social media has been a convenient platform for voicing opinions through posting messages, ranging from tweeting a short text to uploading a media file, or any combination of messages.  ...  More importantly, the joint representation enables emotion-oriented cross-modal retrieval, for example, retrieval of videos using the text query "crazy cat".  ...  In [22] , a ListNet layer is added on top of the RBF layer for ranking the music in valence and arousal in Cartesian coordinates.  ... 
doi:10.1109/tmm.2015.2482228 fatcat:7tozmatnhvbj7hjjohkofngecq

Supervised Quantization for Similarity Search

Xiaojuan Wang, Ting Zhang, Guo-Jun Qi, Jinhui Tang, Jingdong Wang
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, we address the problem of searching for semantically similar images from a large database. We present a compact coding approach, supervised quantization.  ...  The experiments on several standard datasets show the superiority of our approach over the state-of-the art supervised hashing and unsupervised quantization algorithms.  ...  Acknowledgements This work was partially supported by the National Basic Research Program of China (973 Program) under Grant 2014CB347600.  ... 
doi:10.1109/cvpr.2016.222 dblp:conf/cvpr/WangZQTW16 fatcat:j6jk3ellsvhhhes6i6i4rr73pa
« Previous Showing results 1 — 15 out of 538 results