A Metric Learning Reality Check

Publikation: Bidrag til tidsskrift › Konferenceartikel › Forskning › fagfællebedømt

Standard

A Metric Learning Reality Check. / Musgrave, Kevin; Belongie, Serge; Lim, Ser Nam.

I: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2020, s. 681-699.

Publikation: Bidrag til tidsskrift › Konferenceartikel › Forskning › fagfællebedømt

Harvard

Musgrave, K, Belongie, S & Lim, SN 2020, 'A Metric Learning Reality Check', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), s. 681-699. https://doi.org/10.1007/978-3-030-58595-2_41

APA

Musgrave, K., Belongie, S., & Lim, S. N. (2020). A Metric Learning Reality Check. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 681-699. https://doi.org/10.1007/978-3-030-58595-2_41

Vancouver

Musgrave K, Belongie S, Lim SN. A Metric Learning Reality Check. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2020;681-699. https://doi.org/10.1007/978-3-030-58595-2_41

Author

Musgrave, Kevin ; Belongie, Serge ; Lim, Ser Nam. / A Metric Learning Reality Check. I: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2020 ; s. 681-699.

Bibtex

@inproceedings{006b19f527754fad81f6058f115e5f09,

title = "A Metric Learning Reality Check",

abstract = "Deep metric learning papers from the past four years have consistently claimed great advances in accuracy, often more than doubling the performance of decade-old methods. In this paper, we take a closer look at the field to see if this is actually true. We find flaws in the experimental methodology of numerous metric learning papers, and show that the actual improvements over time have been marginal at best. Code is available at github.com/KevinMusgrave/powerful-benchmarker.",

keywords = "Deep metric learning",

author = "Kevin Musgrave and Serge Belongie and Lim, {Ser Nam}",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 16th European Conference on Computer Vision, ECCV 2020 ; Conference date: 23-08-2020 Through 28-08-2020",

year = "2020",

doi = "10.1007/978-3-030-58595-2_41",

language = "English",

pages = "681--699",

journal = "Lecture Notes in Computer Science",

issn = "0302-9743",

publisher = "Springer Verlag",

}

RIS

TY - GEN

T1 - A Metric Learning Reality Check

AU - Musgrave, Kevin

AU - Belongie, Serge

AU - Lim, Ser Nam

PY - 2020

Y1 - 2020

N2 - Deep metric learning papers from the past four years have consistently claimed great advances in accuracy, often more than doubling the performance of decade-old methods. In this paper, we take a closer look at the field to see if this is actually true. We find flaws in the experimental methodology of numerous metric learning papers, and show that the actual improvements over time have been marginal at best. Code is available at github.com/KevinMusgrave/powerful-benchmarker.

AB - Deep metric learning papers from the past four years have consistently claimed great advances in accuracy, often more than doubling the performance of decade-old methods. In this paper, we take a closer look at the field to see if this is actually true. We find flaws in the experimental methodology of numerous metric learning papers, and show that the actual improvements over time have been marginal at best. Code is available at github.com/KevinMusgrave/powerful-benchmarker.

KW - Deep metric learning

UR - http://www.scopus.com/inward/record.url?scp=85097421278&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-58595-2_41

DO - 10.1007/978-3-030-58595-2_41

M3 - Conference article

AN - SCOPUS:85097421278

SP - 681

EP - 699

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

T2 - 16th European Conference on Computer Vision, ECCV 2020

Y2 - 23 August 2020 through 28 August 2020

ER -

ID: 301819335

Datalogisk Institut