3,293 Hits in 7.1 sec

DeepDiffusion: Unsupervised Learning of Retrieval-adapted Representations via Diffusion-based Ranking on Latent Feature Manifold [article]

Takahiko Furuya, Ryutarou Ohbuchi
2021 arXiv   pre-print
Recently proposed neural network-based unsupervised learning approaches have succeeded in obtaining features appropriate for classification of multimedia data.  ...  To obtain such retrieval-adapted features, we introduce the idea of combining diffusion distance on a feature manifold with neural network-based unsupervised feature learning.  ...  Our goal in this paper is to develop an unsupervised learning framework that learns retrievaladapted feature representations and is applicable to a wide range of multimedia data types.  ... 
arXiv:2112.07082v1 fatcat:mc75afshjzdwpmwb5lzguwracy

Unsupervised multi-graph cross-modal hashing for large-scale multimedia retrieval

Liang Xie, Lei Zhu, Guoqi Chen
2016 Multimedia tools and applications  
MGCMH is unsupervised method which integrates multi-graph learning and hash function learning into a joint framework, to learn unified hash space for all modalities.  ...  graph cross-modal hashing for large-scale multimedia retrieval. (2016). Multimedia Tools and Applications. 75, (15), 9185-9204. Research Collection School Of Information Systems.  ...  even be computationally intractable for the retrieval of large-scale multimedia data.  ... 
doi:10.1007/s11042-016-3432-0 fatcat:hn5pwu2l35bqtkjru52c7wfh5i

A hybrid graph-based and non-linear late fusion approach for multimedia retrieval

Ilias Gialampoukidis, Anastasia Moumtzidou, Dimitris Liparas, Stefanos Vrochidis, Ioannis Kompatsiaris
2016 2016 14th International Workshop on Content-Based Multimedia Indexing (CBMI)  
The fusion of multiple modalities for retrieval in an unsupervised way has been mostly based on early, weighted linear, graph-based and diffusion-based techniques.  ...  Nowadays, multimedia retrieval has become a task of high importance, due to the need for efficient and fast access to very large and heterogeneous multimedia collections.  ...  Contrary to the aforementioned approaches for cross-modal retrieval, our proposed framework does not involve a training stage, but proposes an unsupervised fusion of all features.  ... 
doi:10.1109/cbmi.2016.7500252 dblp:conf/cbmi/GialampoukidisM16 fatcat:gggu5spoufduxbycflomf5b7vu

Multimedia indexing and retrieval: ever great challenges

Chabane Djeraba, Moncef Gabbouj, Patrick Bouthemy
2006 Multimedia tools and applications  
In this introduction, we present a brief state of the art of multimedia indexing and retrieval as well as highlight some notions explored in the special issue.  ...  The special issue is actually situated between old problems and new challenges, and contribute to understand the next multimedia indexing and retrieval generation.  ...  We are grateful to the authors for their hard work and patience during this long period of the special issue preparation.  ... 
doi:10.1007/s11042-006-0025-3 fatcat:qlgf7xogufespd3qmv54ssvixu

Semi-automatic audio semantic concept discovery for multimedia retrieval

Yipei Wang, Shourabh Rawat, Florian Metze
2014 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
Previous work explored semantic concepts for content analysis to assist retrieval.  ...  We evaluate the method on NIST 2011 multimedia event detection (MED) dataset.  ...  FRAMEWORK Overview Our proposed method uses an annotation of broad semantic concepts by human as the seed and explores unsupervised method to discover the hidden hierarchical structure of each broad  ... 
doi:10.1109/icassp.2014.6853822 dblp:conf/icassp/WangRM14a fatcat:h5od6ae3njadbbut3czjlavwgi

High diversity transforms multimedia information retrieval into a cross-cutting field

James Z. Wang, Nozha Boujemaa, Yixin Chen
2007 SIGMOD record  
Fan et al. explored an approach to match electronic slides with videos for distance learning. Industrial research papers brought refreshing perspectives to the audience. Jaffe et al. from Yahoo!  ...  Ohbuchi and Kobayashi investigated unsupervised learning of shape features and applied to the retrieval of 3D shape models.  ... 
doi:10.1145/1276301.1276315 fatcat:cdowdmpllbcf3mea7m2wotbuyq

PyRetri: A PyTorch-based Library for Unsupervised Image Retrieval by Deep Convolutional Neural Networks [article]

Benyi Hu, Ren-Jie Song, Xiu-Shen Wei, Yazhou Yao, Xian-Sheng Hua, Yuehu Liu
2020 arXiv   pre-print
In order to fill this gap, we introduce PyRetri, an open source library for deep learning based unsupervised image retrieval.  ...  To the best of our knowledge, this is the first open-source library for unsupervised image retrieval by deep learning.  ...  Figure 1 : 1 Framework of deep learning based unsupervised image retrieval, illustrated with abstractions in PyRetri.  ... 
arXiv:2005.02154v2 fatcat:boghtyahzzfqhftixrzbrvhdri

CNN Retrieval based Unsupervised Metric Learning for Near-Duplicated Video Retrieval [article]

Hao Cheng, Ping Wang, Chun Qi
2021 arXiv   pre-print
Unsupervised metric learning is used for similarity measurement and feature matching.  ...  An efficient re-ranking algorithm combined with k-nearest neighborhood fuses the retrieval results from two levels of features and further improves the retrieval performance.  ...  In respect of feature matching, we leverage an unsupervised metric learning framework to calculate the similarity of videos and kd-tree to store features and do fast retrieval.  ... 
arXiv:2105.14566v1 fatcat:zuxlm3u5j5e3td3axl5fme2gj4

Linear Distance Preserving Pseudo-Supervised and Unsupervised Hashing

Min Wang, Wengang Zhou, Qi Tian, Zhengjun Zha, Houqiang Li
2016 Proceedings of the 2016 ACM on Multimedia Conference - MM '16  
The second one is an unsupervised hashing method, in which quantization loss is considered. We validate our framework on two large-scale datasets.  ...  The experiments demonstrate that our pseudo-supervised method achieves consistent improvement for the state-of-the-art unsupervised hashing methods, while our unsupervised method outperforms the state-of-the-art  ...  This leads to an unsupervised hashing method, we denote it as Linear Distance Transformation Hashing (LDTH).  ... 
doi:10.1145/2964284.2964334 dblp:conf/mm/WangZTZL16 fatcat:xwftms6jybg6jgvkkzuz4wxbia

Deep Self-taught Hashing for Image Retrieval

Ke Zhou, Yu Liu, Jingkuan Song, Linyu Yan, Fuhao Zou, Fumin Shen
2015 Proceedings of the 23rd ACM international conference on Multimedia - MM '15  
deep structures with hashing to improve the retrieval performance by automatically learning robust visual features and hash functions.  ...  Hashing is a promising technique to tackle the problem of scalable retrieval, and it generally consists two major components, namely hash code generation and hash functions learning.  ...  An excellent semantic hash for image retrieval should pre-serve Hamming distance of mapped codes to protect feature similarity from high dimension to low dimension.  ... 
doi:10.1145/2733373.2806320 dblp:conf/mm/ZhouLSYZS15 fatcat:ucrk32nhejh45jd464bi3kcflu

Unsupervised Generative Adversarial Cross-modal Hashing [article]

Jian Zhang, Yuxin Peng, Mingkuan Yuan
2017 arXiv   pre-print
To address the above problem, in this paper we propose an Unsupervised Generative Adversarial Cross-modal Hashing approach (UGACH), which makes full use of GAN's ability for unsupervised representation  ...  helpful to capture meaningful nearest neighbors of different modalities for cross-modal retrieval.  ...  Introduction Multimedia retrieval has become an important application over the past decades, which can retrieve multimedia contents that users have interests in.  ... 
arXiv:1712.00358v1 fatcat:sxe6tprexzfy3h6iknjywwrf7u

Transitive Hashing Network for Heterogeneous Multimedia Retrieval [article]

Zhangjie Cao, Mingsheng Long, Qiang Yang
2016 arXiv   pre-print
domain, which generates transitive hash codes for heterogeneous multimedia retrieval.  ...  Existing work on cross-modal hashing assumes heterogeneous relationship across modalities for hash function learning.  ...  For text network, we employ a three-layer MLP with the numbers of hidden units set to 1000, 500, and b, respectively.  ... 
arXiv:1608.04307v1 fatcat:l2xakppke5b5vgwcgs7wwliufq

Unsupervised learning from a corpus for shape-based 3D model retrieval

Ryutarou Ohbuchi, Jun Kobayashi
2006 Proceedings of the 8th ACM international workshop on Multimedia information retrieval - MIR '06  
In this paper, we explore a method to improve feature distance computation by employing unsupervised learning of the subspace of 3D shape features from a corpus.  ...  The learned manifold is approximated by an RBF network, onto which shape features are projected. Distances among shape features can then be computed effectively on the learned manifold.  ...  Manifold learning has been applied to many problems, including motion analysis, multimedia data retrieval [11] , and others.  ... 
doi:10.1145/1178677.1178701 dblp:conf/mir/OhbuchiK06 fatcat:rtuzoxpij5avlljk4p64lsa4qe

Supervised Hashing with Pseudo Labels for Scalable Multimedia Retrieval

Jingkuan Song, Lianli Gao, Yan Yan, Dongxiang Zhang, Nicu Sebe
2015 Proceedings of the 23rd ACM international conference on Multimedia - MM '15  
There is an increasing interest in using hash codes for efficient multimedia retrieval and data storage.  ...  hash codes can be jointly learned and iteratively updated in an unified framework.  ...  We then prove that the pseudo supervision information and the hash codes can be jointly learned and iteratively updated in an unified framework. • We further integrate the pseudo labels strategy to unsupervised  ... 
doi:10.1145/2733373.2806341 dblp:conf/mm/SongGYZS15 fatcat:xlccv4vb6nd4rf6537agrynpmu

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.  ...  Despite the consistent advances in visual features and other Multimedia Information Retrieval (MIR) techniques, measuring the similarity among multimedia objects is still a challenging task for an effective  ...  The source code is public available, 2 as part of an open-source framework for unsupervised distance learning.  ... 
doi:10.1016/j.patrec.2017.10.013 fatcat:q2gbx7ofgvatrjyxriwtfz7t4y
« Previous Showing results 1 — 15 out of 3,293 results