The Championship-Winning Solution for the 5th CLVISION Challenge 2024

TL;DR

This paper presents a novel approach for the 5th CLVision Challenge 2024, utilizing winning subnetworks, multiple training strategies, and an interaction inference method to handle class recurrence, unlabeled data, and OOD categories, achieving top rankings.

Contribution

The paper introduces a multi-strategy framework with subnetworks and interaction inference to address class recurrence and unlabeled data challenges in incremental learning.

Findings

Achieved first place in both pre-selection and final evaluation stages.

Attained an average accuracy of 0.4535 pre-selection and 0.4805 final.

Effectively handled class recurrence and OOD data in the challenge.

Abstract

In this paper, we introduce our approach to the 5th CLVision Challenge, which presents distinctive challenges beyond traditional class incremental learning. Unlike standard settings, this competition features the recurrence of previously encountered classes and includes unlabeled data that may contain Out-of-Distribution (OOD) categories. Our approach is based on Winning Subnetworks to allocate independent parameter spaces for each task addressing the catastrophic forgetting problem in class incremental learning and employ three training strategies: supervised classification learning, unsupervised contrastive learning, and pseudo-label classification learning to fully utilize the information in both labeled and unlabeled data, enhancing the classification performance of each subnetwork. Furthermore, during the inference stage, we have devised an interaction strategy between subnetworks, where the prediction for a specific class of a particular sample is the average logits across different subnetworks corresponding to that class, leveraging the knowledge learned from different subnetworks on recurring classes to improve classification accuracy. These strategies can be simultaneously applied to the three scenarios of the competition, effectively solving the difficulties in the competition scenarios. Experimentally, our method ranks first in both the pre-selection and final evaluation stages, with an average accuracy of 0.4535 during the preselection stage and an average accuracy of 0.4805 during the final evaluation stage.