XtarNet: Learning to Extract Task-Adaptive Representation for Incremental Few-Shot Learning

By Sung Whan Yoon, Do-Yeon Kim, Jun Seo, Jaekyun Moon. 2020

49 citations recorded by the source corpus.

Learning novel concepts while preserving prior knowledge is a long-standing challenge in machine learning. The challenge gets greater when a novel task is given with only a few labeled examples, a problem known as incremental few-shot learning. We propose XtarNet, which learns to extract task-adaptive representation (TAR) for facilitating incremental few-shot learning. The method utilizes a backbone network pretrained on a set of base categories while also employing additional modules that are meta-trained across episodes. Given a new task, the novel feature extracted from the meta-trained modules is mixed with the base feature obtained from the pretrained model. The process of combining two different features provides TAR and is also controlled by meta-trained modules. The TAR contains effective information for classifying both novel and base categories. The base and novel classifiers quickly adapt to a given task by utilizing the TAR. Experiments on standard image datasets indicate that XtarNet achieves state-of-the-art incremental few-shot learning performance. The concept of TAR can also be used in conjunction with existing incremental few-shot learning methods; extensive simulation results in fact show that applying TAR enhances the known methods significantly.

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

Incremental Few-Shot Semantic Segmentation via Embedding Adaptive-Update and Hyper-class Representation (2022, methodology): “The improvement in segmentation performance with different learning sessions proves that incremental few-shot learning methods (AAN[37] and XtarNet[46]) could reduce catastrophic forgetting by iteratively learning new-class knowledge.”
Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks (2022, background): “There is a similar setting proposed by [71], [72], [73], namely Generalized Few-Shot Learning (GFSL).”
Subspace Regularizers for Few-Shot Class Incremental Learning (2021, methodology): “For testing, base and novel samples have equal weight in average per previous work (Yoon et al., 2020).”
Recent Advances of Continual Learning in Computer Vision: An Overview (2021, positive): “[73] proposed a few-shot continual learning framework called XtarNet containing a base model and different meta-learnable modules.”
Semantic-aware Knowledge Distillation for Few-Shot Class-Incremental Learning (2021, methodology): “[38] proposed a method which obtains a task-adaptive representation for novel tasks based on the information provided from the base task by an attention module.”
Meta Learning on a Sequence of Imbalanced Domains with Difficulty Awareness (2021, background): “Incremental and Continual Few-shot Learning: Incremental few-shot learning [23, 46, 63] aim to learn new categories while retaining knowledge on old categories within a single domain and assume access to the base categories is unlimited.”
Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class Incremental Learning (2023, background): “A popular choice is to use meta learning (Yoon et al., 2020; Chi et al., 2022; Zhou et al., 2022b).”
Learning to Learn and Remember Super Long Multi-Domain Task Sequence (2022, background): “Recent work reveals that catastrophic forgetting often occurs when transferring a meta-learning model to a new context [55, 79].”
Warping the Space: Weight Space Rotation for Class-Incremental Few-Shot Learning (2023, methodology): “To address both catastrophic forgetting and overfitting issues in CIFSL, some prior works take meta-learning approaches (Gidaris & Komodakis, 2018; Ren et al., 2019; Yoon et al., 2020).”
A Survey on Few-Shot Class-Incremental Learning (2023, methodology): “Before the emergence of FSCIL, similar settings had been proposed in previous research, such as those presented by (Gidaris and Ko-modakis, 2018; Qi et al., 2018; Xie et al., 2019; Yoon et al., 2020).”
SemGIF: A Semantics Guided Incremental Few-shot Learning Framework with Generative Replay (2021, methodology): “Table 1 shows the comparison of performance of the proposed framework with the existing literature in both the incremental few-shot [9, 25, 27, 34, 45] and generalised few-shot domains [32, 44].”
Graph Few-shot Class-incremental Learning (2021, methodology): “[31, 46, 47] further extend the situation to the Few-shot Class-incremental Learning (FCL) setting on image domain.”