75,462 Hits in 4.1 sec

Optimization as a Model for Few-Shot Learning

Sachin Ravi, Hugo Larochelle
2017 International Conference on Learning Representations  
We demonstrate that this meta-learning model is competitive with deep metric-learning techniques for few-shot learning.  ...  Here, we propose an LSTMbased meta-learner model to learn the exact optimization algorithm used to train another learner neural network classifier in the few-shot regime.  ...  ACKNOWLEDGMENTS We thank Jake Snell, Kevin Swersky, and Oriol Vinyals for helpful discussions of this work.  ... 
dblp:conf/iclr/RaviL17 fatcat:dq3izbjd7rgjvhil2kf2m5hrkm

Automated Human Cell Classification in Sparse Datasets using Few-Shot Learning [article]

Reece Walsh, Mohamed H. Abdelpakey, Mohamed S. Shehata, Mostafa M.Mohamed
2021 arXiv   pre-print
Third, this study presents future directions for using few-shot learning in human cell classification. In general, few-shot learning in its current state performs poorly on human cell classification.  ...  This study investigates the feasibility of using few-shot learning-based techniques to mitigate the data requirements for accurate training.  ...  Acknowledgements Funding for this project was provided through the MITACS Accelerate grant.  ... 
arXiv:2107.13093v1 fatcat:qqzr6wt75nbgbomnuwgwy2yzti

Insights into few shot learning approaches for image scene classification

Mohamed Soudy, Yasmine Afify, Nagwa Badr
2021 PeerJ Computer Science  
Motivated by these findings, in this paper, we introduce two models for few-shot learning in scene classification.  ...  The first problem is solved by few-shot learning, which achieves optimal performance in object detection and classification but with a lack of eligible attention in the scene classification task.  ...  Meta-learning also incorporates the concept of "learning to learn". The branch of meta-learning known as Few-Shot Learning (FSL) is observing a dramatic increase in research.  ... 
doi:10.7717/peerj-cs.666 pmid:34616882 pmcid:PMC8459776 fatcat:uzwb5qlf75ba3as6jwu4oaoukq

An Introduction to Advanced Machine Learning : Meta Learning Algorithms, Applications and Promises [article]

Farid Ghareh Mohammadi, M. Hadi Amini, Hamid R. Arabnia
2019 arXiv   pre-print
Machine learning employs a strict model or embedded engine to train and predict which still fails to learn unseen classes and sufficiently use online data.  ...  In [1, 2], we have explored the theoretical aspects of feature extraction optimization processes for solving largescale problems and overcoming machine learning limitations.  ...  Gradient decent-based learning This model of MTL is also known as optimization-based model for tuning the parameter (θ ).  ... 
arXiv:1908.09788v1 fatcat:qujten7zzzbd7laazhymnfw2yi

Few-shot learning with adaptively initialized task optimizer: a practical meta-learning approach

Han-Jia Ye, Xiang-Rong Sheng, De-Chuan Zhan
2019 Machine Learning  
By treating the few-shot task as an entirety, extracting task-level pattern, and learning a task-agnostic model initialization, the model-agnostic meta-learning (MAML) framework enables the applications  ...  Directly training a model on such few-shot learning (FSL) tasks falls into the over-fitting dilemma, which would turn to an effective task-level inductive bias as a key supervision.  ...  Acknowledgements This research was supported by the The National Key R&D Program of China (2018YFB1004300), NSFC (61773198, 61751306, 61632004), and the program A for Outstanding Ph.D. candidate of Nanjing  ... 
doi:10.1007/s10994-019-05838-7 fatcat:zu6og3x6yjfwvk5vy3o3g4bogy

Few-Shot Rolling Bearing Fault Diagnosis with Metric-Based Meta Learning

Sihan Wang, Dazhi Wang, Deshan Kong, Jiaxing Wang, Wenhui Li, Shuai Zhou
2020 Sensors  
Therefore, in this paper, we propose a metric-based meta-learning method named Reinforce Relation Network (RRN) for diagnosing bearing faults with few-shot samples.  ...  The comparison studies with other methods demonstrate the superiority of the proposed method in the few-shot scenario.  ...  In few-shot meta learning, only a few samples are given from the training set, for problems such as few-shot scenario and conditions transfer can be effectively solved by few-shot meta learning technology  ... 
doi:10.3390/s20226437 pmid:33187173 fatcat:ilow3gnaavd6jhpkre3nyyxpui

Meta Navigator: Search for a Good Adaptation Policy for Few-shot Learning [article]

Chi Zhang, Henghui Ding, Guosheng Lin, Ruibo Li, Changhu Wang, Chunhua Shen
2021 arXiv   pre-print
Few-shot learning aims to adapt knowledge learned from previous tasks to novel tasks with only a limited amount of labeled data.  ...  learning by seeking a higher-level strategy and proffer to automate the selection from various few-shot learning designs.  ...  Few-shot learning is proposed as a promising direction to alleviate the need for exhaustively labeled data by exploring an extreme case where only a few labeled data is available to undertake a novel task  ... 
arXiv:2109.05749v1 fatcat:f2xay3rt3veijcotnhhpdozpkm

Learning to Generalize to Unseen Tasks with Bilevel Optimization [article]

Hayeon Lee, Donghyun Na, Hae Beom Lee, Sung Ju Hwang
2019 arXiv   pre-print
to be effective for few-shot classification tasks of unseen classes.  ...  To tackle this issue, we propose a simple yet effective meta-learning framework for metricbased approaches, which we refer to as learning to generalize (L2G), that explicitly constrains the learning on  ...  Embedding-based Few-shot Learning Apporaches We briefly describe a generic framework for metric-based few-shot meta-learning methods [8, 13, 5] .  ... 
arXiv:1908.01457v1 fatcat:h4pqh4lcgzg4bgbz7nvq3xfexe

Meta-FSEO: A Meta-Learning Fast Adaptation with Self-Supervised Embedding Optimization for Few-Shot Remote Sensing Scene Classification

Yong Li, Zhenfeng Shao, Xiao Huang, Bowen Cai, Song Peng
2021 Remote Sensing  
In this study, we propose Meta-FSEO, a novel model for improving the performance of few-shot remote sensing scene classification in varying urban scenes.  ...  The proposed Meta-FSEO model deploys self-supervised embedding optimization for adaptive generalization in new tasks such as classifying features in new urban regions that have never been encountered during  ...  Acknowledgments: The authors are sincerely grateful to Steve McClure for revised the grammatical errors in the paper, and the editors, as well as the anonymous reviewers, for their valuable suggestions  ... 
doi:10.3390/rs13142776 fatcat:mf32i6f2wrbnfmpkdda3j3m3ny

Uncertainty-Aware Few-Shot Image Classification [article]

Zhizheng Zhang, Cuiling Lan, Wenjun Zeng, Zhibo Chen, Shih-Fu Chang
2021 arXiv   pre-print
In this work, we propose Uncertainty-Aware Few-Shot framework for image classification by modeling uncertainty of the similarities of query-support pairs and performing uncertainty-aware optimization.  ...  Few-shot image classification learns to recognize new categories from limited labelled data.  ...  For few-shot learning classification, the "hardness" of a few-shot episode is quantified with a metric in (Dhillon et al.  ... 
arXiv:2010.04525v2 fatcat:fjfzjvrrurf3xfp2mz4czsmlwq

MetAdapt: Meta-Learned Task-Adaptive Architecture for Few-Shot Classification [article]

Sivan Doveh, Eli Schwartz, Chao Xue, Rogerio Feris, Alex Bronstein, Raja Giryes, Leonid Karlinsky
2020 arXiv   pre-print
Few-Shot Learning (FSL) is a topic of rapidly growing interest.  ...  These modules are added to the model and are meta-trained to predict the optimal network connections for a given novel task.  ...  In [1] models are trained to perform set-operations (e.g. union) and then can be used to synthesise samples for few-shot multi-label classifications. Few-shot meta-learning (learning-to-learn).  ... 
arXiv:1912.00412v3 fatcat:2zy6imvmfndrtfqzcvtm2zj6z4

MCML: A Novel Memory-based Contrastive Meta-Learning Method for Few Shot Slot Tagging [article]

Hongru Wang, Zezhong Wang, Gabriel Pui Cheong Fung, Kam-Fai Wong
2021 arXiv   pre-print
Meta-learning is widely used for few-shot slot tagging in the task of few-shot learning. The performance of existing methods is, however, seriously affected by catastrophic forgetting.  ...  Experimental results show that MCML is scalable and outperforms metric-based meta-learning and optimization-based meta-learning on all 1shot, 5-shot, 10-shot, and 20-shot scenarios of the SNIPS dataset  ...  Related Work Few-Shot Learning Few-shot learning is first proposed as a transfer method using a Bayesian approach on low-level visual features (Li Fei-Fei, Fergus, and Perona, 2006) .  ... 
arXiv:2108.11635v2 fatcat:km4sqqk3wzbnxivuf6l22ohwna

A Hybrid Approach with Optimization and Metric-based Meta-Learner for Few-Shot Learning [article]

Duo Wang, Yu Cheng, Mo Yu, Xiaoxiao Guo, Tao Zhang
2019 arXiv   pre-print
Few-shot learning aims to learn classifiers for new classes with only a few training examples per class.  ...  Our meta-metric-learning approach consists of two components, a task-specific metric-based learner as a base model, and a meta-learner that learns and specifies the base model.  ...  A Hybrid Approach with Optimization and Metric-based Meta-Learner for Few-Shot Learning Duo Wang, Yu Cheng, Mo Yu, Xiaoxiao Guo, and Tao Zhang, Senior Member, IEEE Abstract-Few-shot learning aims to learn  ... 
arXiv:1904.03014v2 fatcat:lkdqydb5e5dyrmx5u5j7f4btoa

Meta-learning via Language Model In-context Tuning [article]

Yanda Chen, Ruiqi Zhong, Sheng Zha, George Karypis, He He
2022 arXiv   pre-print
The goal of meta-learning is to learn to adapt to a new task with only a few labeled examples.  ...  , the labeled examples, and the target input to predict; to meta-train the model to learn from in-context examples, we fine-tune a pre-trained language model (LM) to predict the target label from the input  ...  As expected, directly optimizing the few-shot in-context learning objective (Section 2.2) improves the fewshot in-context learning accuracy. Few-shot examples lead to more effective task adaptation.  ... 
arXiv:2110.07814v2 fatcat:6izoe7b23fht7iov4k5doxkm54

Few-shot Action Recognition with Implicit Temporal Alignment and Pair Similarity Optimization [article]

Congqi Cao, Yajuan Li, Qinyi Lv, Peng Wang, Yanning Zhang
2020 arXiv   pre-print
similarity comparison; 3) an advanced loss for few-shot learning to optimize pair similarity with limited data.  ...  Specifically, we propose a novel few-shot action recognition framework that uses long short-term memory following 3D convolutional layers for sequence modeling and alignment.  ...  Pair Similarity Optimization Module Few-shot learning can be seen as maximizing intra-class similarity as well as minimizing inter-class similarity.  ... 
arXiv:2010.06215v1 fatcat:hhchjwegsjfkxnpv4xoykxy5vu
« Previous Showing results 1 — 15 out of 75,462 results