Boosting Few-Shot Visual Learning With Self-Supervision

By Spyros Gidaris, Andrei Bursuc, Nikos Komodakis, Patrick Pérez, Matthieu Cord. 2019

453 citations recorded by the source corpus.

Few-shot learning and self-supervised learning address different facets of the same problem: how to train a model with little or no labeled data. Few-shot learning aims for optimization methods and models that can learn efficiently to recognize patterns in the low data regime. Self-supervised learning focuses instead on unlabeled data and looks into it for the supervisory signal to feed high capacity deep neural networks. In this work we exploit the complementarity of these two domains and propose an approach for improving few-shot learning through self-supervision. We use self-supervision as an auxiliary task in a few-shot learning pipeline, enabling feature extractors to learn richer and more transferable visual representations while still using few annotated samples. Through self-supervision, our approach can be naturally extended towards using diverse unlabeled data from other datasets in the few-shot setting. We report consistent improvements across an array of architectures, datasets and self-supervision techniques. We provide the implementation code at: https://github.com/valeoai/BF3S

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Representation & Self-Supervised Learning

In Meta & Continual Learning

Citations in the corpus

CrossTransformers: spatially-aware few-shot transfer (2020, background): “Some recent works [29, 77] demonstrate that this can improve few-shot learning, although these use self-supervised auxiliary losses rather than integrating self-supervised instance discrimination [15, 24, 55, 81, 94] into episodic training.”
S3C: Self-Supervised Stochastic Classifiers for Few-Shot Class-Incremental Learning (2023, methodology): “These features can notably improve the performance of downstream tasks like few-shot learning [13], semi-supervised learning [43], to improve the model robustness [14], class imbalance [40], etc. Recently, Lee et al. [22] used SSL to improve the performance for supervised classification, by…”
Few-Shot and Continual Learning with Attentive Independent Mechanisms (2021, methodology): “In few-shot learning [12, 37, 45, 14], the goal is to learn novel concepts with as few samples as possible, i.”
Few-Shot Lifelong Learning (2021, methodology): “We perform experiments with the auxiliary task as relative patch location prediction (Doersch, Gupta, and Efros 2015) (Patch), rotation angle prediction (Rotation) (Gidaris et al. 2019) and SimCLR (Chen et al. 2020).”
Incremental Few-Shot Object Detection (2020, neutral): “Note that there are many more recent FSL methods [6, 26, 7, 17, 47, 55, 50, 27, 39, 11, 16] that produce better performance on image classification tasks than MAML.”
When Does Self-supervision Improve Few-shot Learning? (2019, background): “Our results also echo those of Gidaris et al. (2019) who ﬁnd that the rotation task improves on mini - and tiered -ImageNet.”
Curriculum-Based Meta-learning (2021, background): “A plenty of ProtoNet variants were proposed for improving model generalization ability[11, 21, 22, 35, 45].”
Compositional Few-Shot Recognition with Primitive Discovery and Enhancing (2020, methodology): “Therefore, inspired by current works [19, 31, 37] on self-supervised learning, we propose to use the split-based self-supervised mechanism for FSL, where we split the input image horizontally and vertically, perm the splits, and ask the model to recognize which perms are applied to them.”
A Survey on Self-Supervised Learning: Algorithms, Applications, and Future Trends (2023, unknown): “…[145], multi-task learning [157], [158], transfer learning [159], [160], [161], graph NNs [147], [162], [163], reinforcement learning [164], [165], [166], few-shot learning [167], [168], neural architecture search [169], robust learning [146], [170], [171], [172], and meta-learning [173], [174].”
Self-Supervised CycleGAN for Object-Preserving Image-to-Image Domain Adaptation (2020, background): “Gidaris et al. [7] integrated the self-supervised learning task to the few-shot learning framework for exploiting richer and more transferable visual representations from few annotated samples.”
Empirical Bayes Transductive Meta-Learning with Synthetic Gradients (2020, methodology): “For 5-shot learning, SIB is sligtly worse than CTM (Li et al., 2019) and/or Gidaris et al. (2019).”
A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities (2022, methodology): “33% BF3S(patch location) [127] Conv-4-64 Multitask Learning 54.”
Few-Shot Learning With a Strong Teacher (2021, background): “Auxiliary self-supervised tasks are also incorporated to improve the generalization ability of the resulting few-shot learners or classiﬁers [79], [80], [81].”
Striking a Balance between Stability and Plasticity for Class-Incremental Learning (2021, methodology): “3(c), we use two residual BasicBlocks [13, 8] and a cosine classiﬁer to map embeddings to a latent space for rotation prediction.”
On Episodes, Prototypical Networks, and Few-shot Learning (2020, methodology): “We chose to analyse Prototypical Networks not only for their simplicity, but also because they often appear as important building blocks of newly-proposed methods (e.g. Oreshkin et al. (2018); Cao et al. (2020); Gidaris et al. (2019); Yoon et al. (2019)).”
Learning Representations by Predicting Bags of Visual Words (2020, methodology): “Moreover, recent work has shown that self-supervision can be beneficial to many other learning problems [10, 18, 28, 29, 58, 71], such as few-shot [18, 58] and semi-supervised [28, 71] learning, or training generative adversarial networks [10].”
Meta-Learning with Shared Amortized Variational Inference (2020, methodology): “Cosine similarity based classifiers have recently been widely adopted in few-shot classification (Dvornik et al., 2019; Gidaris et al., 2019; Lee et al., 2019; Oreshkin et al., 2018; Ye et al., 2018).”