Synthesized Feature based Few-Shot Class-Incremental Learning on a Mixture of Subspaces

By Ali Cheraghian, Shafin Rahman, Sameera Ramasinghe, Pengfei Fang, Christian Simon, Lars Petersson, Mehrtash Harandi. 2021

81 citations recorded by the source corpus.

Few-shot class incremental learning (FSCIL) aims to incrementally add sets of novel classes to a well-trained base model in multiple training sessions with the restriction that only a few novel instances are available per class. While learning novel classes, FSCIL methods gradually forget base (old) class training and overfit to a few novel class samples. Existing approaches have addressed this problem by computing the class prototypes from the visual or semantic word vector domain. In this paper, we propose addressing this problem using a mixture of subspaces. Subspaces define the cluster structure of the visual domain and help to describe the visual and semantic domain considering the overall distribution of the data. Additionally, we propose to employ a variational autoencoder (VAE) to generate synthesized visual samples for augmenting pseudo-feature while learning novel classes incrementally. The combined effect of the mixture of subspaces and synthesized features reduces the forgetting and overfitting problem of FSCIL. Extensive experiments on three image classification datasets show that our proposed method achieves competitive results compared to state-of-the-art methods.

Subfields: Artificial Intelligence, Cancer Research, 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 Cancer Research

In Computer Vision and Pattern Recognition

In Representation & Self-Supervised Learning

In Meta & Continual Learning

Citations in the corpus

A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “[71], Mixture of Subspaces [72], Subspace Reg [12], FSIL-”
S3C: Self-Supervised Stochastic Classifiers for Few-Shot Class-Incremental Learning (2023, background): “The works proposed in [9,8,1,10,48,21,32] focus on learning robust manifolds by regular-izing feature space representations.”
Few-Shot Class-Incremental Learning from an Open-Set Perspective (2022, neutral): “We compare our method with the SOTA methods [17,1,11,19,4,5,23] on three datasets.”
Few-Shot Class-Incremental Learning via Entropy-Regularized Data-Free Replay (2022, methodology): “The comparisons with the most recent method [8] illustrate the superiority of our proposed one.”
Constrained Few-shot Class-incremental Learning (2022, neutral): “Session (s) 1 2 3 4 5 6 7 8 9 No. of classes |C̃(s)| 60 65 70 75 80 85 90 95 100 Mode/ Work Feature extractor AL-MML [48] ResNet-18 64.10 55.88 47.07 45.16 40.11 36.38 33.96 31.55 29.37 Semantic KD∗ [6] ResNet-18 <64 <57 <51 <46 <43 <41 <39 <37 <35 VAE∗ [7] ResNet-18 <62 <58 <57 <52 <51 <49 <46 <45 ”
Warping the Space: Weight Space Rotation for Class-Incremental Few-Shot Learning (2023, background): “There have been many attempts to solve class-incremental few-shot learning where only a few labeled samples are available in each task (Cheraghian et al., 2021b; Zhu et al., 2021; Zhou et al., 2022b; Cheraghian et al., 2021a).”
MetaFSCIL: A Meta-Learning Approach for Few-Shot Class Incremental Learning (2022, background): “[5] Cheraghian, Ali and Rahman, Shafin and Ramasinghe, Sameera and Fang, Pengfei and Simon, Christian and Petersson, Lars and Harandi, Mehrtash.”
A Survey on Few-Shot Class-Incremental Learning (2023, methodology): “In Cheraghian et al. (2021b), a mixture of subspaces is proposed to describe the visual and semantic domain distribution of the data, which helps to avoid forgetting old classes.”
Neural Collapse Inspired Feature-Classifier Alignment for Few-Shot Class Incremental Learning (2023, background): “Some studies try to make base and incremental sessions compatible via pseudo-feature (Cheraghian et al., 2021b; Zhou et al., 2022a), augmentation (Peng et al., 2022), or looking for a ﬂat minima (Shi et al., 2021).”
SIN: Semantic Inference Network for Few-Shot Streaming Label Learning (2022, background): “A more practical setting, incremental few-shot learning (IFSL) [17], [44]–[47], is recently proposed with the goal of undertaking the multi-class incremental learning task with limited data in each incremental step.”