Filters








1,348 Hits in 3.2 sec

Online Supervised Hashing for Ever-Growing Datasets [article]

Fatih Cakir, Sarah Adel Bargal, Stan Sclaroff
2015 arXiv   pre-print
Supervised hashing methods are widely-used for nearest neighbor search in computer vision applications. Most state-of-the-art supervised hashing approaches employ batch-learners.  ...  To address these issues, we propose an online hashing method that is amenable to changes and expansions of the datasets.  ...  Using this approach they report state-of-the-art performance for supervised hashing.  ... 
arXiv:1511.03257v1 fatcat:vadjkkvlvrhl5nx6t4mcmt7t6a

Online supervised hashing

Fatih Cakir, Stan Sclaroff
2015 2015 IEEE International Conference on Image Processing (ICIP)  
Fast similarity search is becoming more and more critical given the ever growing sizes of datasets.  ...  However, most supervised hashing methods are batch-learners; this hinders their ability to adapt to changes as a dataset grows and diversifies.  ...  To our knowledge, this is the first supervised hashing method that allows the label space to grow.  ... 
doi:10.1109/icip.2015.7351274 dblp:conf/icip/CakirS15 fatcat:2lrqldude5ddphozbscy5vxcvu

Online supervised hashing

Fatih Cakir, Sarah Adel Bargal, Stan Sclaroff
2017 Computer Vision and Image Understanding  
Fast similarity search is becoming more and more critical given the ever growing sizes of datasets.  ...  However, most supervised hashing methods are batch-learners; this hinders their ability to adapt to changes as a dataset grows and diversifies.  ...  To our knowledge, this is the first supervised hashing method that allows the label space to grow.  ... 
doi:10.1016/j.cviu.2016.10.009 fatcat:x5z2nohcwbfr7omtipm6hamvn4

Online Hashing for Scalable Remote Sensing Image Retrieval

Peng Li, Xiaoyu Zhang, Xiaobin Zhu, Peng Ren
2018 Remote Sensing  
Therefore, the pre-trained hash functions might not fit the ever-growing new RS images.  ...  Supervised hashing approaches, such as kernel-based supervised hashing [25], supervised discrete hashing [27] and deep hashing methods [29], incorporate the label information to learn semantic hashing  ...  Our proposed online hashing method can be used in many real-time remote sensing applications due to its adapting ability to variations in datasets as they grow and diversify.  ... 
doi:10.3390/rs10050709 fatcat:2dbb6rhsb5agha4ve5e6t7dmj4

Online Hashing with Similarity Learning [article]

Zhenyu Weng, Yuesheng Zhu
2021 arXiv   pre-print
The experiments on two multi-label image datasets show that our method is competitive or outperforms the state-of-the-art online hashing methods in terms of both accuracy and efficiency for multi-label  ...  However, when the hash functions are updated, the binary codes for the whole database have to be updated to be consistent with the hash functions, resulting in the inefficiency in the online image retrieval  ...  Although these methods can achieve good search performance, they are designed for the static database and have poor scalability for the ever-growing database.  ... 
arXiv:2108.02560v1 fatcat:gordxlzmvbeghlo2nw5gh6gsdi

Hadamard Matrix Guided Online Hashing [article]

Mingbao Lin, Rongrong Ji, Hong Liu, Xiaoshuai Sun, Shen Chen, Qi Tian
2020 arXiv   pre-print
To handle the above challenges, a novel supervised online hashing scheme termed Hadamard Matrix Guided Online Hashing (HMOH) is proposed in this paper.  ...  Online image hashing has attracted increasing research attention recently, which receives large-scale data in a streaming manner to update the hash functions on-the-fly.  ...  Existing works in online hashing can be categorized into either supervised methods or unsupervised methods.  ... 
arXiv:1905.04454v3 fatcat:mzw4soe6lbd5xfn2dd6omqveju

Online Hashing with Efficient Updating of Binary Codes

Zhenyu Weng, Yuesheng Zhu
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Online hashing methods are efficient in learning the hash functions from the streaming data.  ...  However, when the hash functions change, the binary codes for the database have to be recomputed to guarantee the retrieval accuracy.  ...  accumulating the ever-increasing database inevitably blocks the timeliness of the online retrieval process.  ... 
doi:10.1609/aaai.v34i07.6920 fatcat:zjgwbir6dvb5bkgj37kknqsnli

Metagenomic binning through low density hashing [article]

Yunan Luo, Y. William Yu, Jianyang Zeng, Bonnie Berger, Jian Peng
2017 bioRxiv   pre-print
Here, we present Opal for metagenomic binning, the task of identifying the origin species of DNA sequencing reads.  ...  Our tool is up to two orders of magnitude faster than leading alignment-based methods at similar or improved accuracy, allowing computational tractability on large metagenomic datasets.  ...  Code Availability Source code for Opal can be found online at http://opal.csail.mit.edu, and through the linked Github repository. All code has been published under the GNU General Public License.  ... 
doi:10.1101/133116 fatcat:nnbcxflhl5e27lgsdy24w3etcy

Submodular video hashing

Liangliang Cao, Zhenguo Li, Yadong Mu, Shih-Fu Chang
2012 Proceedings of the 20th ACM international conference on Multimedia - MM '12  
For a larger scale synthetic dataset with 1M samples, it uses less than 1 second in response to 100 queries.  ...  In the hashing stage, we represent each video component as a compact hash code, and combine multiple hash codes into hash tables for effective search.  ...  The U.S. government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright thereon.  ... 
doi:10.1145/2393347.2393393 dblp:conf/mm/CaoLMC12 fatcat:tfd6cuhxo5crpnoii3olapxn6m

Online Hashing with Efficient Updating of Binary Codes [article]

Zhenyu Weng, Yuesheng Zhu
2019 arXiv   pre-print
Online hashing methods are efficient in learning the hash functions from the streaming data.  ...  However, when the hash functions change, the binary codes for the database have to be recomputed to guarantee the retrieval accuracy.  ...  accumulating the ever-increasing database inevitably blocks the timeliness of the online retrieval process.  ... 
arXiv:1911.12125v2 fatcat:z56eu2oznfhz5ejnpocwusvyyu

CaPSuLe: A camera-based positioning system using learning

Yongshik Moon, Soonhyun Noh, Daedong Park, Chen Luo, Anshumali Shrivastava, Seongsoo Hong, Krishna Palem
2016 2016 29th IEEE International System-on-Chip Conference (SOCC)  
Such a significant gain in computation and energy cost is a result of careful choices of hash tables, hash functions, and related operations.  ...  2100 Joules of energy for getting the current location; entirely impractical for use in a mobile context.  ...  The response time and energy consumption for bruteforce and our approach are evaluated. 1) Accuracy: For our dataset, the accuracy of bruteforce is 93%, Bow 75%, supervised learning 77% and CaPSuLe 92.11%  ... 
doi:10.1109/socc.2016.7905476 dblp:conf/socc/MoonNPLSHP16 fatcat:pqdmd3qfczg5ne7g2a6n5eywse

Visual Vocabulary Learning and Its Application to 3D and Mobile Visual Search [article]

Liujuan Cao
2012 arXiv   pre-print
Especial focuses would be also given for the recent trends in supervised/unsupervised vocabulary optimization, compact descriptor for visual search, as well as in multi-view based 3D object representation  ...  The State of The Arts The ever growing computational power motivates the research efforts to extract visual descriptors directly on a mobile device [20, 21, 23, 116] - [121] .  ...  Coming with the ever growing computational power in the mobile devices, recent works have proposed to directly extract compact visual descriptors on the mobile devices [20, 21, 116] - [118] .  ... 
arXiv:1207.7244v1 fatcat:b6y7yvvcu5davkwh3zuv762qq4

Learning Better Encoding for Approximate Nearest Neighbor Search with Dictionary Annealing [article]

Shicong Liu, Hongtao Lu
2015 arXiv   pre-print
with residual vectors, "cools down" the dictionary by fitting the intermediate dataset, then extracts the new residual vectors for the next iteration.  ...  An optimal series of dictionaries should be mutually independent, and each dictionary should generate a balanced encoding for the target dataset. Existing methods did not explicitly consider this.  ...  Experiments show that our online dictionary learning substantially further improves the ANN search quality, which makes vector quantization methods more effective to the ever-growing dataset in the real  ... 
arXiv:1507.01442v1 fatcat:64deguaupbh2lci4ppvvdjnqmu

Image and video mining through online learning

Andrew Gilbert, Richard Bowden
2017 Computer Vision and Image Understanding  
However, the labelling of the training set is time-consuming, especially as datasets grow in size and complexity.  ...  By repeating this process in an online learning framework, the accuracy of similarity increases dramatically despite labelling only a few training examples.  ...  We also provide analysis regarding cluster purity and evaluation of the computational cost of the approach, showing that the online learning framework can compete favourably with the state of the art supervised  ... 
doi:10.1016/j.cviu.2017.02.001 fatcat:ieatxhrqzffljlnn4ao43sgwcy

Play and Rewind

Hanwang Zhang, Meng Wang, Richang Hong, Tat-Seng Chua
2016 Proceedings of the 2016 ACM on Multimedia Conference - MM '16  
In this paper, we propose a novel unsupervised video hashing framework called Self-Supervised Temporal Hashing (SSTH) that is able to capture the temporal nature of videos in an end-to-end learning-to-hash  ...  The hash function is learned in a self-supervised fashion, where a decoder RNN is proposed to reconstruct the original video frames in both forward and reverse orders.  ...  Without a doubt, the ever-growing abundance of videos on the Web has brought about an urgent need for more advanced CBVR technologies [18, 39, 40, 37] . Video is beyond a set of frames.  ... 
doi:10.1145/2964284.2964308 dblp:conf/mm/ZhangWHC16 fatcat:jmuptaw46jht3ipuau7qrvqsry
« Previous Showing results 1 — 15 out of 1,348 results