Embedding Propagation: Smoother Manifold for Few-Shot Classification

Rodríguez, Paul; Laradji, Issam; Drouin, Alexandre; Lacoste, Alexandre

doi:10.1007/978-3-030-58574-7_8

Cited by 147 publications

(97 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1. Metric learning methods (i.e., MatchingNets, 113 ProtoNets, 114 RelationNets, 115 Graph neural network (GraphNN), 116 Ridge regression, 117 TransductiveProp, 118 Fine-tuning Baseline, 119 URT, 120 DSN-MR, 121 CDFS, 122 DeepEMD, 123 EPNet, 124 ACC + Amphibian, 125 FEAT, 126 T A B L E 3 (Continued)…”

Section: Discussion About Different Meta-learningsmentioning

confidence: 99%

“…We can divide meta‐learning methods into three categories 140 : Metric learning methods (i.e., MatchingNets, 113 ProtoNets, 114 RelationNets, 115 Graph neural network (GraphNN), 116 Ridge regression, 117 TransductiveProp, 118 Fine‐tuning Baseline, 119 URT, 120 DSN‐MR, 121 CDFS, 122 DeepEMD, 123 EPNet, 124 ACC + Amphibian, 125 FEAT, 126 MsSoSN + SS + SD + DD, 127 RFS, 128 RFS + CRAT, 129 IDA, 130 LR + ICI, 131 FEAT + MLMT, 132 BOHB, 133 CSPN, 134 SUR, 135 SKD, 136 TAFSSL, 137 TRPN, 138 and TransMatch 139 ) learn a similarity space in which learning is particularly efficient for few‐shot examples. Memory network methods (i.e., Meta Networks, 103 TADAM, 104 MCFS, 105 and MRN 106 ) learn to store “experience” when learning seen tasks and then generalize it to unseen tasks. …”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Fighting fire with fire: A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification

et al. 2021

View full text Add to dashboard Cite

Adversarial images generated by generative adversarial networks are not close to any existing benign images, and contain nonrobust features that have been identified as critical to the robustness of a machine learning model. Since adversarial images have an underlying distribution that differs from normal images, these kinds of images can offer valuable features for training a robust model. To deal with these special features, we focus on a novel machine learning task of adversarial images classification, where adversarial images can be used to investigate the problem of classifying adversarial images themselves. In the setting of this novel task, adversarial images are the ONLY kind of data used in training and testing, rather than not just a set of testing images as usual. To this end, we propose a novel spatial–frequency ensemble relation network with generative adversarial learning. First, we present a spatial–frequency ensemble representation learning to extract the feature of training images. Second, we design a meta‐learning‐based relation model to gain the relationship between images. Third, to achieve a robust model, we utilize generative adversarial learning and transform the relationship into a Jacobian matrix. Finally, we design a discriminator model that determines whether an adversarial image is from the matching category or not. Experimental results demonstrate that our approach achieves significantly higher performance compared with other state‐of‐the‐arts.

show abstract

Section: Discussion About Different Meta-learningsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Fighting fire with fire: A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification

et al. 2021

View full text Add to dashboard Cite

show abstract

“…A benchmark on miniImageNet [52] and tieredImageNet [53] shows superior performance compared to other state-of-the-art FSL algorithms especially using zero to five shots, which means it works especially well if only few labels are available. However, a typical issue with FSL is that the training and test samples are disjoint [18]. This causes the feature extractor of a TPN to produce embeddings that are seemingly uncorrelated for unseen classes.…”

Section: Missing Labelsmentioning

confidence: 99%

“…This manifests as a disadvantage in terms of robustness when the TPN tries to propagate the labels during graph construction. The Embedding Propagation Network (EPNet) [18] addresses this shortcoming of TPNs by applying the propagation at embedding creation time, thus locating an image's embedding close to images with similar features in embedding space, resulting in closer labels in their respective space. EPNet achieves superior performance over the TPN architecture in one-and five-shot benchmarking.…”

Section: Missing Labelsmentioning

confidence: 99%

A Survey of Un-, Weakly-, and Semi-Supervised Learning Methods for Noisy, Missing and Partial Labels in Industrial Vision Applications

Simmler

Sager²,

Andermatt³

et al. 2021

2021 8th Swiss Conference on Data Science (SDS)

View full text Add to dashboard Cite

When applying deep learning methods in an industrial vision application, they often fall short of the performance shown in a clean and controlled lab environment due to data quality issues. Few would consider the actual labels as a driving factor, yet inaccurate label data can impair model performance significantly. However, being able to mitigate inaccurate or incomplete labels might also be a cost-saver for real-world projects.Here, we survey state-of-the-art deep learning approaches to resolve such missing labels, noisy labels, and partially labeled data in the prospect of an industrial vision application. We systematically present un-, weakly, and semi-supervised approaches from 'A' like anomaly detection to 'Z' like zero-shot classification to resolve these challenges by embracing them.

show abstract

“…These methods learn a support vector based on the feature representation of training images in order to create a cluster for each label; queries are labeled based on the label of their nearby cluster. Label propagation [97] and embedding propagation [131] are also other techniques researchers are using to address the few-shot learning problem.…”

Section: Experiments and Resultsmentioning

confidence: 99%

Deep learning based monocular visual-inertial odometry

Esfahani¹

View full text Add to dashboard Cite

Embedding Propagation: Smoother Manifold for Few-Shot Classification

Cited by 147 publications

References 27 publications

Fighting fire with fire: A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification

Fighting fire with fire: A spatial–frequency ensemble relation network with generative adversarial learning for adversarial image classification

A Survey of Un-, Weakly-, and Semi-Supervised Learning Methods for Noisy, Missing and Partial Labels in Industrial Vision Applications

Deep learning based monocular visual-inertial odometry

Contact Info

Product

Resources

About