Learning to Propagate Labels: Transductive Propagation Network for Few-Shot Learning

By Yanbin Liu, Juho Lee, Minseop Park, Saehoon Kim, Eunho Yang, Sung Ju Hwang, Yi Yang. 2018

758 citations recorded by the source corpus.

The goal of few-shot learning is to learn a classifier that generalizes well even when trained with a limited number of training instances per class. The recently introduced meta-learning approaches tackle this problem by learning a generic classifier across a large number of multiclass classification tasks and generalizing the model to a new task. Yet, even with such meta-learning, the low-data problem in the novel classification task still remains. In this paper, we propose Transductive Propagation Network (TPN), a novel meta-learning framework for transductive inference that classifies the entire test set at once to alleviate the low-data problem. Specifically, we propose to learn to propagate labels from labeled instances to unlabeled test instances, by learning a graph construction module that exploits the manifold structure in the data. TPN jointly learns both the parameters of feature embedding and the graph construction in an end-to-end manner. We validate TPN on multiple benchmark datasets, on which it largely outperforms existing few-shot learning approaches and achieves the state-of-the-art results.

Subfields: Artificial Intelligence

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

Citations in the corpus

Masked Label Prediction: Unified Massage Passing Model for Semi-Supervised Classification (2020, methodology): “APPNP [ Klicpera et al. , 2018 ] and TPN [ Liu et al. , 2019 ] propose to use GCN to predict soft labels and then propagate them with Personalized Pagerank.”
Meta-Learning with Less Forgetting on Large-Scale Non-Stationary Task Distributions (2022, positive): “It has been shown that existing works [47,35,36] are particularly sensitive to OOD data.”
XtarNet: Learning to Extract Task-Adaptive Representation for Incremental Few-Shot Learning (2020, neutral): “66% Transductive Propagation Nets (Liu et al., 2018) Conv4 55.”
An Ensemble of Epoch-Wise Empirical Bayes for Few-Shot Learning (2019, background): “rity space, in which the learning should be efﬁcient for fewshot examples. The metrics include Euclidean distance [32], cosine distance [7], [36], relation module [33] and graphbased similarity [37], [38]. Metric-based task-speciﬁc feature representation learning has also been presented in many rela”
Few-Shot Learning via Embedding Adaptation With Set-to-Set Functions (2018, neutral): “We compare with two previous approaches, TPN [25] and TEAM [29].”
On Modulating the Gradient for Meta-learning (2020, background): “Note that, our meta-learning approach is an adaptive module acting on the model parameters which diﬀers from relation and metric learning approaches e.g ., [17,23,35,36,42,43].”
Relational Generalized Few-Shot Learning (2019, background): “Alternatively, GCNs may exploit the manifold structure in the data to propagate labels from labeled to unlabeled images by using edge weights that depend on learned distances in the feature space [37].”
Part-aware Prototype Network for Few-shot Semantic Segmentation (2020, background): “Most of existing works can be categorized into metriclearning based [34,28,32], optimization-learning based [23,8], and graph neural network [9,18] based methods.”
Adaptive Risk Minimization: A Meta-Learning Approach for Tackling Group Shift (2020, background): “, 2019) or the transductive learning setting (Liu et al., 2019; Antoniou & Storkey, 2019; Hu et al., 2020).”
Unifying Graph Convolutional Neural Networks and Label Propagation (2020, methodology): “…Algorithm (LPA) (Zhu et al., 2005; Zhou et al., 2004; Zhang & Lee, 2007; Wang & Zhang, 2008; Karasuyama & Mamitsuka, 2013; Gong et al., 2017; Liu et al., 2019a); On the node side V , node features are propagated along edges and transformed through neural network layers, which is known as…”
Hyperbolic Image Embeddings (2019, neutral): “71% TPN [25] ResNet12 59.”
SimpleShot: Revisiting Nearest-Neighbor Classification for Few-Shot Learning (2019, neutral): “: Results reported in [19].”
Cross Attention Network for Few-shot Classification (2019, methodology): “Following [26, 17, 36, 21], we use ResNet-12 network as our embedding module.”