Multi-Level Contrastive Learning for Few-Shot Problems

Contrastive learning is a discriminative approach that aims at grouping similar samples closer and diverse samples far from each other. It it an efficient technique to train an encoder generating distinguishable and informative representations, and it may even increase the encoder's transferability. Most current applications of contrastive learning benefit only a single representation from the last layer of an encoder.In this paper, we propose a multi-level contrasitive learning approach which applies contrastive losses at different layers of an encoder to learn multiple representations from the encoder. Afterward, an ensemble can be constructed to take advantage of the multiple representations for the downstream tasks. We evaluated the proposed method on few-shot learning problems and conducted experiments using the mini-ImageNet and the tiered-ImageNet datasets. Our model achieved the new state-of-the-art results for both datasets, comparing to previous regular, ensemble, and contrastive learing (single-level) based approaches.

[1]  Myriam Tami,et al.  Spatial Contrastive Learning for Few-Shot Classification , 2020, ECML/PKDD.

[2]  Oriol Vinyals,et al.  Matching Networks for One Shot Learning , 2016, NIPS.

[3]  Thomas L. Griffiths,et al.  Recasting Gradient-Based Meta-Learning as Hierarchical Bayes , 2018, ICLR.

[4]  Subhransu Maji,et al.  Meta-Learning With Differentiable Convex Optimization , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Geoffrey E. Hinton,et al.  A Simple Framework for Contrastive Learning of Visual Representations , 2020, ICML.

[6]  Bernt Schiele,et al.  An Ensemble of Epoch-Wise Empirical Bayes for Few-Shot Learning , 2019, ECCV.

[7]  Bernt Schiele,et al.  Learning to Self-Train for Semi-Supervised Few-Shot Classification , 2019, NeurIPS.

[8]  Bernt Schiele,et al.  Meta-Transfer Learning for Few-Shot Learning , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  Sindy Löwe,et al.  Putting An End to End-to-End: Gradient-Isolated Learning of Representations , 2019, NeurIPS.

[10]  Joshua Achiam,et al.  On First-Order Meta-Learning Algorithms , 2018, ArXiv.

[11]  Ling Yang,et al.  DPGN: Distribution Propagation Graph Network for Few-Shot Learning , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Guizhong Liu,et al.  Few-Shot Image Classification via Contrastive Self-Supervised Learning , 2020, ArXiv.

[13]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[14]  Seungjin Choi,et al.  Gradient-Based Meta-Learning with Learned Layerwise Metric and Subspace , 2018, ICML.

[15]  Cordelia Schmid,et al.  What makes for good views for contrastive learning , 2020, NeurIPS.

[16]  Hoirin Kim,et al.  Transductive Few-shot Learning with Meta-Learned Confidence , 2020, ArXiv.

[17]  Xiaogang Wang,et al.  Finding Task-Relevant Features for Few-Shot Learning by Category Traversal , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[18]  Vincent Gripon,et al.  Exploiting Unsupervised Inputs for Accurate Few-Shot Classification , 2020, ArXiv.

[19]  Alexandre Lacoste,et al.  TADAM: Task dependent adaptive metric for improved few-shot learning , 2018, NeurIPS.

[20]  Cordelia Schmid,et al.  Diversity With Cooperation: Ensemble Methods for Few-Shot Classification , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[21]  Kaiming He,et al.  Improved Baselines with Momentum Contrastive Learning , 2020, ArXiv.

[22]  Joshua B. Tenenbaum,et al.  Meta-Learning for Semi-Supervised Few-Shot Classification , 2018, ICLR.

[23]  Fei Sha,et al.  Few-Shot Learning via Embedding Adaptation With Set-to-Set Functions , 2018, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[24]  Tao Xiang,et al.  Learning to Compare: Relation Network for Few-Shot Learning , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[25]  Razvan Pascanu,et al.  Meta-Learning with Latent Embedding Optimization , 2018, ICLR.

[26]  Richard S. Zemel,et al.  Prototypical Networks for Few-shot Learning , 2017, NIPS.

[27]  Xilin Chen,et al.  Cross Attention Network for Few-shot Classification , 2019, NeurIPS.

[28]  Leonid Karlinsky,et al.  TAFSSL: Task-Adaptive Feature Sub-Space Learning for few-shot classification , 2020, ECCV.

[29]  Orchid Majumder,et al.  Revisiting Contrastive Learning for Few-Shot Classification , 2021, ArXiv.

[30]  Phillip Isola,et al.  Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere , 2020, ICML.

[31]  Alexandre Drouin,et al.  Embedding Propagation: Smoother Manifold for Few-Shot Classification , 2020, ECCV.

[32]  R. Devon,et al.  Putting An End to End-to-End: Gradient-Isolated Learning of Representations , 2019 .

[33]  Andrew Zisserman,et al.  Automatically Discovering and Learning New Visual Categories with Ranking Statistics , 2020, ICLR.

[34]  Nakamasa Inoue,et al.  Semi-Supervised Contrastive Learning with Generalized Contrastive Loss and Its Application to Speaker Recognition , 2020, 2020 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC).