Filters








242,190 Hits in 4.8 sec

Deep Metric Learning to Rank

Fatih Cakir, Kun He, Xide Xia, Brian Kulis, Stan Sclaroff
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
We propose a novel deep metric learning method by revisiting the learning to rank approach.  ...  To fully exploit the benefits of the ranking formulation, we also propose a new minibatch sampling scheme, as well as a simple heuristic to enable large-batch training.  ...  We revisit learning to rank for deep metric learning, and propose to learn a distance metric by optimizing Average Precision (AP) [4] over entire ranked lists.  ... 
doi:10.1109/cvpr.2019.00196 dblp:conf/cvpr/Cakir0XKS19 fatcat:dp7zjz36ovhtlgqhe4zgjchvhm

Learning to Rank Words: Optimizing Ranking Metrics for Word Spotting [article]

Pau Riba, Adrià Molina, Lluis Gomez, Oriol Ramos-Terrades, Josep Lladós
2021 arXiv   pre-print
In this paper, we explore and evaluate the use of ranking-based objective functions for learning simultaneously a word string and a word image encoder.  ...  We consider retrieval frameworks in which the user expects a retrieval list ranked according to a defined relevance score.  ...  Ranking metrics In word spotting and, more generally in information retrieval, several metrics have been carefully designed to evaluate the obtained rankings.  ... 
arXiv:2106.05144v1 fatcat:m7hmrwkpbjhdxdvj2wvujwt7ma

A General Framework for Counterfactual Learning-to-Rank [article]

Aman Agarwal, Kenta Takatsu, Ivan Zaitsev, Thorsten Joachims
2019 arXiv   pre-print
., click, dwell time) is an attractive source of training data for Learning-to-Rank, but its naive use leads to learning results that are distorted by presentation bias.  ...  Going beyond this special case, this paper provides a general and theoretically rigorous framework for counterfactual learning-to-rank that enables unbiased training for a broad class of additive ranking  ...  UNBIASED LEARNING FOR RANK-BASED IR METRICS We begin by generalizing the counterfactual learning framework from [11] to the class of linearly decomposable metrics as defined below.  ... 
arXiv:1805.00065v3 fatcat:cjvtma4rrjglrmlgckcn7uh5gu

Learning to Rank Using Localized Geometric Mean Metrics

Yuxin Su, Irwin King, Michael Lyu
2017 Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval - SIGIR '17  
Based on the combination of local learned metrics, we employ the popular Normalized Discounted Cumulative Gain (NDCG) scorer and Weighted Approximate Rank Pairwise (WARP) loss to optimize the ideal candidate  ...  The experiments on real-world datasets demonstrate that our proposed L-GMML algorithm outperforms the state-of-the-art metric learning to rank methods and the stylish query-independent LtR algorithms regarding  ...  ACKNOWLEDGMENTS We would like to thank the anonymous reviewers for their valuable comments and suggestions to improve the quality of the paper.  ... 
doi:10.1145/3077136.3080828 dblp:conf/sigir/SuKL17 fatcat:aju2b5ymyzb2rapxv5i6gfcid4

Learning to rank in person re-identification with metric ensembles [article]

Sakrapee Paisitkriangkrai, Chunhua Shen, Anton van den Hengel
2015 arXiv   pre-print
More importantly, we advance state-of-the-art results on person re-identification by improving the rank-1 recognition rates from 40% to 50% on the iLIDS benchmark, 16% to 18% on the PRID2011 benchmark,  ...  We propose an effective structured learning based approach to the problem of person re-identification which outperforms the current state-of-the-art on most benchmark data sets evaluated.  ...  LMNN is later applied to a task of person re-identification in [14] . Wu et al. applies the Metric Learning to Rank (MLR) method of [23] to person re-id [37] .  ... 
arXiv:1503.01543v1 fatcat:v6nuk7jaabhu5mf6rjqw2jf34u

Learning to Combine Multiple Ranking Metrics for Fault Localization

Jifeng Xuan, Martin Monperrus
2014 2014 IEEE International Conference on Software Maintenance and Evolution  
In this paper, we propose MULTRIC, a learning-based approach to combining multiple ranking metrics for effective fault localization.  ...  Spectrum-based fault localization applies a ranking metric to identify faulty source code.  ...  In contrast to above work [27] , our MULTRIC is a combination of existing ranking metrics via learning-to-rank techniques.  ... 
doi:10.1109/icsme.2014.41 dblp:conf/icsm/XuanM14 fatcat:macyauptlvfv3i56ihca74rr4a

Ranking journals: could Google Scholar Metrics be an alternative to Journal Citation Reports and Scimago Journal Rank?

Emilio Delgado-López-Cózar, Álvaro Cabezas-Clavijo
2013 Learned Publishing  
The launch of Google Scholar Metrics as a tool for assessing scientific journals may be serious competition for Thomson Reuters Journal Citation Reports, and for Scopus powered Scimago Journal Rank.  ...  Despite its shortcomings, Google Scholar Metrics is a helpful tool for authors and editors in identifying core journals.  ...  The authors would like to thank Nicolás Robinson-García for translating this text and Alberto Martín-Martín for journals data processing  ... 
doi:10.1087/20130206 fatcat:uy3wbgmjffe77ecbdh3dpks2ia

Learning to rank in person re-identification with metric ensembles

Sakrapee Paisitkriangkrai, Chunhua Shen, Anton van den Hengel
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
More importantly, we advance state-of-the-art results on person re-identification by improving the rank-1 recognition rates from 40% to 50% on the iLIDS benchmark, 16% to 18% on the PRID2011 benchmark,  ...  We propose an effective structured learning based approach to the problem of person re-identification which outperforms the current state-of-the-art on most benchmark data sets evaluated.  ...  LMNN is later applied to a task of person re-identification in [14] . Wu et al. applies the Metric Learning to Rank (MLR) method of [25] to person re-id [43] .  ... 
doi:10.1109/cvpr.2015.7298794 dblp:conf/cvpr/Paisitkriangkrai15 fatcat:ijxrgydssvb2rp422lvedhbkqq

Rank-based distance metric learning: An application to image retrieval

Jung-Eun Lee, Rong Jin, Anil K. Jain
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
We present a novel approach to learn distance metric for information retrieval.  ...  To address this problem explicitly, we propose rankbased distance metric learning.  ...  We aim to address this problem by a rank-based distance metric learning.  ... 
doi:10.1109/cvpr.2008.4587389 dblp:conf/cvpr/LeeJJ08 fatcat:bo7gdygwuvgjrdfxrxtc4gcgqe

On the suitability of diversity metrics for learning-to-rank for diversity

Rodrygo L.T. Santos, Craig Macdonald, Iadh Ounis
2011 Proceedings of the 34th international ACM SIGIR conference on Research and development in Information - SIGIR '11  
Moreover, the suitability of these metrics is compromised as they try to penalise redundancy during the learning process.  ...  In this paper, we contrast the suitability of relevance and diversity metrics as objective functions for learning a diverse ranking.  ...  metrics for learning a diverse ranking.  ... 
doi:10.1145/2009916.2010111 dblp:conf/sigir/SantosMO11a fatcat:drxbbtuogvbjha7hlzekeoafum

Ranking with semi-supervised distance metric learning and its application to housing potential estimation

Yangqiu Song, Bin Zhang, Wenjun Yin, Changshui Zhang, Jin Dong
2007 Proceedings of the sixteenth ACM conference on Conference on information and knowledge management - CIKM '07  
This paper proposes a semi-supervised distance metric learning algorithm for the ranking problem.  ...  Then the computer can automatically use the most certain points and plenty of unlabeded data to learn an informative metric for ranking.  ...  For ranking, learning an informative distance metric is helpful to reduce the computational complexity and improve the ranking accuracy.  ... 
doi:10.1145/1321440.1321589 dblp:conf/cikm/SongZYZD07 fatcat:2dxrt6u6engtdd7my3gnpljwnm

Automatic Image Captions for Lightly Labelled Images

Raju Janagam, K. Yakub Reddy
2018 International Journal of Trend in Scientific Research and Development  
We initiate a new distance metric learning technique recognized as ambiguously supervised structural metric learning to find out discriminative Mahalanobis distance metric that is based on weak supervision  ...  Here we introduce two methods to correspondingly get hold of two discriminative affinity matrices by means of learning from the images of weakly labelled.  ...  within metric learning to rank.  ... 
doi:10.31142/ijtsrd10786 fatcat:aglkknrdpjabbc34lj7vm6sf5a

Relevance Ranking Metrics for Learning Objects

X. Ochoa, E. Duval
2008 IEEE Transactions on Learning Technologies  
Moreover, combining the metrics through learning algorithms sorts the result list 50 percent better than the baseline ranking.  ...  The main objective of this paper is to improve the current status of learning object search.  ...  ACKNOWLEDGMENTS This work was supported by the cooperation agreement between FWO (Belgium) and Senacyt (Ecuador) under the project "Smart Tools to Find and Reuse Learning Objects."  ... 
doi:10.1109/tlt.2008.1 fatcat:kiuwpvxwebaipdwwwr7pxowila

Online Low-Rank Metric Learning via Parallel Coordinate Descent Method

Gan Sun, Yang Cong, Qiang Wang, Xiaowei Xu
2018 2018 24th International Conference on Pattern Recognition (ICPR)  
trace norm to promote low-rankness on the transformation matrix.  ...  In order to tackle these challenges, in this paper, we intend to establish a robust metric learning formulation with the expectation that online metric learning and parallel optimization can solve large-scale  ...  ONLINE LOW-RANK METRIC LEARNING A.  ... 
doi:10.1109/icpr.2018.8546239 dblp:conf/icpr/SunCWX18 fatcat:zfdkrte72nb2te3icqh5wv2hka

Scalable Metric Learning via Weighted Approximate Rank Component Analysis [article]

Cijo Jose, Francois Fleuret
2016 arXiv   pre-print
We propose a metric learning formulation called Weighted Approximate Rank Component Analysis (WARCA).  ...  WARCA optimizes the precision at top ranks by combining the WARP loss with a regularizer that favors orthonormal linear mappings, and avoids rank-deficient embeddings.  ...  Information theoretic metric learning [25] (ITML) exploits the relationship between the Mahalanobis distance and Gaussian distributions to learn a metric by minimizing the KL-divergence with a metric  ... 
arXiv:1603.00370v2 fatcat:diaqkjwup5fwjk4laywxkjenqi
« Previous Showing results 1 — 15 out of 242,190 results