Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ...
Deep Metric Learning (DML) is a group of techniques that aim to measure the similarity between objects through the neural network. Although the number of DML methods has rapidly increased in recent years, most previous studies cannot effectively handle noisy data, which commonly exists in practical...
Main Authors: | , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
Underline Science Inc.
2023
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48448/qmrb-0p03 https://underline.io/lecture/69179-denoising-multi-similarity-formulation-a-self-paced-curriculum-driven-approach-for-robust-metric-learning |
id |
ftdatacite:10.48448/qmrb-0p03 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48448/qmrb-0p03 2023-06-11T04:11:18+02:00 Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... Association for Artificial Intelligence 2023 Gu, Bin Luo, Lei Zhang, Chenkang 2023 https://dx.doi.org/10.48448/qmrb-0p03 https://underline.io/lecture/69179-denoising-multi-similarity-formulation-a-self-paced-curriculum-driven-approach-for-robust-metric-learning unknown Underline Science Inc. Computer and Information Science Machine Learning Artificial Intelligence Audiovisual MediaObject Conference talk article 2023 ftdatacite https://doi.org/10.48448/qmrb-0p03 2023-06-01T11:26:55Z Deep Metric Learning (DML) is a group of techniques that aim to measure the similarity between objects through the neural network. Although the number of DML methods has rapidly increased in recent years, most previous studies cannot effectively handle noisy data, which commonly exists in practical applications and often leads to serious performance deterioration. To overcome this limitation, in this paper, we build a connection between noisy samples and hard samples in the framework of self-paced learning, and propose a \underline{B}alanced \underline{S}elf-\underline{P}aced \underline{M}etric \underline{L}earning (BSPML) algorithm with a denoising multi-similarity formulation, where noisy samples are treated as extremely hard samples and adaptively excluded from the model training by sample weighting. Especially, due to the pairwise relationship and a new balance regularization term, the sub-problem \emph{w.r.t.} sample weights is a nonconvex quadratic function. To efficiently solve this nonconvex ... Article in Journal/Newspaper DML DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Computer and Information Science Machine Learning Artificial Intelligence |
spellingShingle |
Computer and Information Science Machine Learning Artificial Intelligence Association for Artificial Intelligence 2023 Gu, Bin Luo, Lei Zhang, Chenkang Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... |
topic_facet |
Computer and Information Science Machine Learning Artificial Intelligence |
description |
Deep Metric Learning (DML) is a group of techniques that aim to measure the similarity between objects through the neural network. Although the number of DML methods has rapidly increased in recent years, most previous studies cannot effectively handle noisy data, which commonly exists in practical applications and often leads to serious performance deterioration. To overcome this limitation, in this paper, we build a connection between noisy samples and hard samples in the framework of self-paced learning, and propose a \underline{B}alanced \underline{S}elf-\underline{P}aced \underline{M}etric \underline{L}earning (BSPML) algorithm with a denoising multi-similarity formulation, where noisy samples are treated as extremely hard samples and adaptively excluded from the model training by sample weighting. Especially, due to the pairwise relationship and a new balance regularization term, the sub-problem \emph{w.r.t.} sample weights is a nonconvex quadratic function. To efficiently solve this nonconvex ... |
format |
Article in Journal/Newspaper |
author |
Association for Artificial Intelligence 2023 Gu, Bin Luo, Lei Zhang, Chenkang |
author_facet |
Association for Artificial Intelligence 2023 Gu, Bin Luo, Lei Zhang, Chenkang |
author_sort |
Association for Artificial Intelligence 2023 |
title |
Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... |
title_short |
Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... |
title_full |
Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... |
title_fullStr |
Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... |
title_full_unstemmed |
Denoising Multi-Similarity Formulation: A Self-paced Curriculum-Driven Approach for Robust Metric Learning ... |
title_sort |
denoising multi-similarity formulation: a self-paced curriculum-driven approach for robust metric learning ... |
publisher |
Underline Science Inc. |
publishDate |
2023 |
url |
https://dx.doi.org/10.48448/qmrb-0p03 https://underline.io/lecture/69179-denoising-multi-similarity-formulation-a-self-paced-curriculum-driven-approach-for-robust-metric-learning |
genre |
DML |
genre_facet |
DML |
op_doi |
https://doi.org/10.48448/qmrb-0p03 |
_version_ |
1768386237615308800 |