Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ...
The extraction of useful deep features is important for many computer vision tasks. Deep features extracted from classification networks have proved to perform well in those tasks. To obtain features of greater usefulness, end-to-end distance metric learning (DML) has been applied to train the featu...
Main Authors: | , , |
---|---|
Format: | Text |
Language: | unknown |
Published: |
arXiv
2017
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.1712.10151 https://arxiv.org/abs/1712.10151 |
id |
ftdatacite:10.48550/arxiv.1712.10151 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.1712.10151 2023-06-11T04:11:17+02:00 Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... Horiguchi, Shota Ikami, Daiki Aizawa, Kiyoharu 2017 https://dx.doi.org/10.48550/arxiv.1712.10151 https://arxiv.org/abs/1712.10151 unknown arXiv https://dx.doi.org/10.1109/tpami.2019.2911075 arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Computer Vision and Pattern Recognition cs.CV FOS Computer and information sciences article-journal ScholarlyArticle Text Article 2017 ftdatacite https://doi.org/10.48550/arxiv.1712.1015110.1109/tpami.2019.2911075 2023-06-01T12:09:36Z The extraction of useful deep features is important for many computer vision tasks. Deep features extracted from classification networks have proved to perform well in those tasks. To obtain features of greater usefulness, end-to-end distance metric learning (DML) has been applied to train the feature extractor directly. However, in these DML studies, there were no equitable comparisons between features extracted from a DML-based network and those from a softmax-based network. In this paper, by presenting objective comparisons between these two approaches under the same network architecture, we show that the softmax-based features perform competitive, or even better, to the state-of-the-art DML features when the size of the dataset, that is, the number of training samples per class, is large. The results suggest that softmax-based features should be properly taken into account when evaluating the performance of deep features. ... : 6 pages ... Text 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 Vision and Pattern Recognition cs.CV FOS Computer and information sciences |
spellingShingle |
Computer Vision and Pattern Recognition cs.CV FOS Computer and information sciences Horiguchi, Shota Ikami, Daiki Aizawa, Kiyoharu Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... |
topic_facet |
Computer Vision and Pattern Recognition cs.CV FOS Computer and information sciences |
description |
The extraction of useful deep features is important for many computer vision tasks. Deep features extracted from classification networks have proved to perform well in those tasks. To obtain features of greater usefulness, end-to-end distance metric learning (DML) has been applied to train the feature extractor directly. However, in these DML studies, there were no equitable comparisons between features extracted from a DML-based network and those from a softmax-based network. In this paper, by presenting objective comparisons between these two approaches under the same network architecture, we show that the softmax-based features perform competitive, or even better, to the state-of-the-art DML features when the size of the dataset, that is, the number of training samples per class, is large. The results suggest that softmax-based features should be properly taken into account when evaluating the performance of deep features. ... : 6 pages ... |
format |
Text |
author |
Horiguchi, Shota Ikami, Daiki Aizawa, Kiyoharu |
author_facet |
Horiguchi, Shota Ikami, Daiki Aizawa, Kiyoharu |
author_sort |
Horiguchi, Shota |
title |
Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... |
title_short |
Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... |
title_full |
Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... |
title_fullStr |
Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... |
title_full_unstemmed |
Significance of Softmax-based Features in Comparison to Distance Metric Learning-based Features ... |
title_sort |
significance of softmax-based features in comparison to distance metric learning-based features ... |
publisher |
arXiv |
publishDate |
2017 |
url |
https://dx.doi.org/10.48550/arxiv.1712.10151 https://arxiv.org/abs/1712.10151 |
genre |
DML |
genre_facet |
DML |
op_relation |
https://dx.doi.org/10.1109/tpami.2019.2911075 |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.1712.1015110.1109/tpami.2019.2911075 |
_version_ |
1768386225080631296 |