Learning Prompt with Distribution-Based Feature Replay for Few-Shot Class-Incremental Learning

TL;DR

This paper introduces LP-DiF, a novel framework leveraging vision-language models and distribution-based feature replay to improve few-shot class-incremental learning, achieving state-of-the-art results on multiple benchmarks.

Contribution

The paper proposes a new FSCIL method using CLIP and prompt tuning with pseudo-feature replay based on Gaussian distributions, enhancing old knowledge retention and new knowledge acquisition.

Findings

Achieves SOTA performance on CIFAR100, mini-ImageNet, CUB-200, SUN-397.

Effectively prevents forgetting through pseudo-feature replay.

Leverages vision-language models for improved FSCIL.

Abstract

Few-shot Class-Incremental Learning (FSCIL) aims to continuously learn new classes based on very limited training data without forgetting the old ones encountered. Existing studies solely relied on pure visual networks, while in this paper we solved FSCIL by leveraging the Vision-Language model (e.g., CLIP) and propose a simple yet effective framework, named Learning Prompt with Distribution-based Feature Replay (LP-DiF). We observe that simply using CLIP for zero-shot evaluation can substantially outperform the most influential methods. Then, prompt tuning technique is involved to further improve its adaptation ability, allowing the model to continually capture specific knowledge from each session. To prevent the learnable prompt from forgetting old knowledge in the new session, we propose a pseudo-feature replay approach. Specifically, we preserve the old knowledge of each class by maintaining a feature-level Gaussian distribution with a diagonal covariance matrix, which is estimated by the image features of training images and synthesized features generated from a VAE. When progressing to a new session, pseudo-features are sampled from old-class distributions combined with training images of the current session to optimize the prompt, thus enabling the model to learn new knowledge while retaining old knowledge. Experiments on three prevalent benchmarks, i.e., CIFAR100, mini-ImageNet, CUB-200, and two more challenging benchmarks, i.e., SUN-397 and CUB-200$^*$ proposed in this paper showcase the superiority of LP-DiF, achieving new state-of-the-art (SOTA) in FSCIL. Code is publicly available at https://github.com/1170300714/LP-DiF.