Infinite Mixture Prototypes for Few-Shot Learning Adaptively - PowerPoint PPT Presentation

Nov 14, 2023 •247 likes •369 views

Infinite Mixture Prototypes for Few-Shot Learning Adaptively inferring model capacity for simple and complex tasks Kelsey Allen, Evan Shelhamer, Hanul Shin, Josh Tenenbaum Few-Shot Learning by Deep Metric Learning Given few instances of a few

Infinite Mixture Prototypes for Few-Shot Learning Adaptively inferring model capacity for simple and complex tasks Kelsey Allen, Evan Shelhamer*, Hanul Shin*, Josh Tenenbaum
Few-Shot Learning by Deep Metric Learning Given few instances of a few classes, recognize a new instance: Labeled support Query
Few-Shot Learning by Deep Metric Learning Given few instances of a few classes, recognize a new instance: Labeled support Query deep net
Few-Shot Learning by Deep Metric Learning Given few instances of a few classes, recognize a new instance: Labeled support Query embedding embedding deep net
Few-Shot Learning by Deep Metric Learning Given few instances of a few classes, recognize a new instance: Unlabeled Labeled support support Query embedding embedding deep net
Few-Shot Learning by Deep Metric Learning Given few instances of a few classes, recognize a new instance: Unlabeled Labeled support support Query embedding embedding deep net deep net
Few-Shot Learning by Deep Metric Learning Given few instances of a few classes, recognize a new instance: Unlabeled Labeled support support Query embedding embedding deep net deep net
Simple and Complex Tasks ● Simple tasks might be accurately represented as uni-modal clusters ● Complex tasks might require a more sophisticated clustering ● A deeper/wider network may not solve both kinds of task simultaneously Omniglot super category task Omniglot character task
Simple and Complex Tasks ● Simple tasks might be accurately represented as uni-modal clusters ● Complex tasks might require a more sophisticated clustering ● A deeper/wider network may not solve both kinds of task simultaneously Omniglot character embeddings Omniglot super category embeddings
Infinite Mixture Modeling ● Represent clustering process using Dirichlet Process mixture model ● Unbounded number of clusters in mixture - let data determine for itself ● Naturally interpolates between nearest neighbors (each data point its own cluster) and prototypes (each cluster is uni-modal Gaussian) ● Semi-supervised and unsupervised possible
Adaptive Capacity for Simple and Complex Tasks ● Adapt between simple and complex data distributions by learning deep representation and inferring the number of clusters ● Efficient inference based on DP-means
Results Poster 87 - 25% absolute improvement over prototypical nets (Snell et al. 2017) for alphabet/super-class recognition on Omniglot - 10% absolute improvement for super-class to sub-class transfer on tiered-ImageNet - equal or better to fully-supervised and semi-supervised prototypical nets on Omniglot and mini-ImageNet benchmarks - 7% absolute improvement over deep nearest neighbors on mini-ImageNet - 20% absolute improvement in unsupervised clustering AMI

Recommend

Zero-Shot Learning for Word Translation: Successes and Failures Ndapa Nakashole, University of

Zero-Shot Learning for Word Translation: Successes and Failures Ndapa Nakashole, University of California, San Diego 05 June 2018 Outline Introduction Successes Limitations 2 Zero-shot learning Zero-shot learning: ) at test

881 views • 49 slides

Co-Representation Network for Generalized Zero-Shot Learning Fei Zhang, Guangming Shi XIDIAN

Co-Representation Network for Generalized Zero-Shot Learning Fei Zhang, Guangming Shi XIDIAN UNIVERSITY ICML 2019 In Intr troduct oduction ion Classic Deep CNN Data requirements decrease Predict Transfer Learning Few-Shot

456 views • 11 slides

Siamese Network & Matching Network for one-shot learning Reference Papers Siamese Neural

Reading Group 2016.11.22 Siamese Network & Matching Network for one-shot learning Reference Papers Siamese Neural Networks for One-Shot Image Recognition (Gregory Koch, Ruslan Salakhutdinov) Matching Network for One-shot Learning (Oriol

802 views • 16 slides

Convergence of latent mixing measures in finite and infinite mixture models Long Nguyen

Convergence of latent mixing measures in finite and infinite mixture models Long Nguyen Department of Statistics University of Michigan BNP Workshop, ICERM 2012 Nguyen@BNP (ICERM12) 1 / 29 Outline Identifiability and consistency in

509 views • 36 slides

Stochastic Processes, Kernel Regression, Infinite Mixture Models Gabriel Huang (TA for Simon

Stochastic Processes, Kernel Regression, Infinite Mixture Models Gabriel Huang (TA for Simon Lacoste-Julien) IFT 6269 : Probabilistic Graphical Models - Fall 2018 Stochastic Process = Random Function 2 Today Motivate Gaussian and

1.13k views • 79 slides

Infinite Positive Semidefinite Tensor Factorization for Source Separation of Mixture Signals

Infinite Positive Semidefinite Tensor Factorization by K. Yoshii, R. Tomioka, D. Mochihashi, and M. Goto Infinite Positive Semidefinite Tensor Factorization for Source Separation of Mixture Signals Kazuyoshi Yoshii Ryota Tomioka Daichi

461 views • 8 slides

Outline Motivation and challenge Dirichlet Process and Infinite Mixture Formulation

Nonparametric Bayesian M Nonparametric Bayesian Models odels --Learning and Reasoning in Open Possible Worlds -- Learning and Reasoning in Open Possible Worlds Eric Xing epxing@cs.cmu.edu Machine Learning Dept./Language Technology

783 views • 40 slides

Gaussian Mixture Models & EM CE-717: Machine Learning Sharif University of Technology M.

Gaussian Mixture Models & EM CE-717: Machine Learning Sharif University of Technology M. Soleymani Fall 2016 Mixture Models: definition Mixture models: Linear supper-position of mixtures or components | =

1.04k views • 39 slides

Concepts with Few-shot Supervision Xuming He ShanghaiTech University

Learning Structured Visual Concepts with Few-shot Supervision Xuming He ShanghaiTech University hexm@shanghaitech.edu.cn 1 12/5/2019 Outline Introduction Learning from very limited annotated data Background in few-shot

807 views • 52 slides

Foundations of Machine Learning Learning with Infinite Hypothesis Sets Motivation With an

Foundations of Machine Learning Learning with Infinite Hypothesis Sets Motivation With an infinite hypothesis set , the error bounds H of the previous lecture are not informative. Is efficient learning from a finite sample possible when

1.24k views • 42 slides

Laplacian Regularized Few Shot Learning (LaplacianShot) Imtiaz Masud Ziko, Jose Dolz, Eric

Laplacian Regularized Few Shot Learning (LaplacianShot) Imtiaz Masud Ziko, Jose Dolz, Eric Granger and Ismail Ben Ayed ETS Montreal 1 Overview Few-Shot Proposed Experiments Learning LaplacianShot - Experimental Setup - What and Why ?

1.22k views • 55 slides

Meta-transfer Learning for Few-shot Learning Yaoyao Liu Tianjin University and NUS School of

NUS-Tsinghua-Southampton Centre for Extreme Search Meta-transfer Learning for Few-shot Learning Yaoyao Liu Tianjin University and NUS School of Computing OUTLINE Research Background Methods Meta-transfer Learning Hard-task

455 views • 33 slides

EM Algorithm and Mixture Models Guojun Zhang University of Waterloo Unsupervised learning and

EM Algorithm and Mixture Models Guojun Zhang University of Waterloo Unsupervised learning and clustering Learn the intrinsic representation of unlabeled data Other examples: density estimation, novelty detection Mixture model

218 views • 9 slides

Non-Parametric Few-Shot Learning CS 330 1 Logistics Homework 1 due tonight, Homework 2 out soon

Non-Parametric Few-Shot Learning CS 330 1 Logistics Homework 1 due tonight, Homework 2 out soon Fill out project group form if you havent already. Project suggestions & project spreadsheet posted 2 Plan for Today Non-Parametric Few-Shot

676 views • 34 slides

Federated Zero-Shot Learning: A Proposal Francesco Odierna CS PhD student @ University of Pisa

Federated Zero-Shot Learning: A Proposal Francesco Odierna CS PhD student @ University of Pisa francesco.odierna@phd.unipi.it 1 Outline Introduction Federated Learning Zero-Shot Learning The Proposal Motivation Road

349 views • 19 slides

Applied Machine Learning Expectation Maximization for Mixture of Gaussians Siamak Ravanbakhsh

Applied Machine Learning Expectation Maximization for Mixture of Gaussians Siamak Ravanbakhsh COMP 551 (Fall 2020) Learning objectives what is a latent variable model? Gaussian mixture model the intuition behind the Expectation-Maximization

530 views • 18 slides

AND MACHINE LEARNING CHAPTER 10: MIXTURE MODELS AND EM Mixture Models - Define a joint

PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 10: MIXTURE MODELS AND EM Mixture Models - Define a joint distribution over observed and latent variables - The corresponding distribution of the observed variables alone is obtained by

878 views • 62 slides

Bag-of-components: an online algorithm for batch learning of mixture models Olivier Schwander

Information Geometry for mixtures Co-Mixture Models Bag of components Bag-of-components: an online algorithm for batch learning of mixture models Olivier Schwander Frank Nielsen Universit Pierre et Marie Curie, Paris, France cole

551 views • 20 slides

Machine Learning Fall 2017 Unsupervised Learning (Clustering: k -means, EM, mixture models)

Machine Learning Fall 2017 Unsupervised Learning (Clustering: k -means, EM, mixture models) Professor Liang Huang (Chaps. 15-16 of CIML) Roadmap CIML Chaps. 3, 4,5,7,11,17 ,18 so far: (large-margin) supervised learning online learning:

815 views • 54 slides

Embedding and Data Augmentation yanweifu@fudan.edu.cn

One-shot Learning in Semantic Embedding and Data Augmentation yanweifu@fudan.edu.cn http://yanweifu.github.io One-shot Learning: learning object categories from just a few images, by incorporating

724 views • 61 slides

Optimization-Based Meta-Learning ( fi nishing from last time) and Non-Parametric Few-Shot

Optimization-Based Meta-Learning ( fi nishing from last time) and Non-Parametric Few-Shot Learning CS 330 1 Logistics Homework 1 due, Homework 2 out this Wednesday Fill out poster presentation preferences ! (Tues 12/3 or Weds 12/4) Course

554 views • 30 slides

Few-shot learning of weak supervision sources in Snorkel (or, learning weakly supervised weak

Few-shot learning of weak supervision sources in Snorkel (or, learning weakly supervised weak supervisors) Jesse Mu Project Outline Replicate Snorkel causal relation extraction system Learn weak supervision sources from tiny sets of

364 views • 12 slides

Prototypes and Matrix Relevance Learning in Complex Fourier Space M. Straat, M. Kaden, M. Gay, T.

Prototypes and Matrix Relevance Learning in Complex Fourier Space M. Straat, M. Kaden, M. Gay, T. Villmann, A. Lampe, U. Seiffert, M. Biehl, and F. Melchert June 26, 2017 Overview A study of classification of time series. In Fourier space:

470 views • 23 slides

Nonlinear Stein Variational Gradient Descent for Learning Diversified Mixture Models Dilin Wang

Nonlinear Stein Variational Gradient Descent for Learning Diversified Mixture Models Dilin Wang Qiang Liu Department of Computer Science The University of Texas at Austin Dilin Wang and Qiang Liu Nonlinear SVGD 1 / 8 Learning Mixture

306 views • 9 slides