Official implementation for IEEE 2025 Paper "GeNGuide: Generalized Neighbour Guidance Framework for Noisy Class Incremental Learning"
This work addresses the challenging, real-world problem of Noisy-label Class Incremental Learning, where label noise adversely affects the performance at each incremental task. Towards this goal, we propose a two-stage approach leveraging pre-trained models effectively and information from neighbour- ing samples. In the first stage, we introduce a neighbour- guidance cross-entropy loss function for feature learning, where the contributions of each data sample is computed based on the labels of neighbouring samples in the feature space, which also evolves as training progresses. This adaptive loss prioritizes reliable samples and mitigates the influence of noisy ones. In the second stage, classifiers are refined using weighted class statistics, which is also guided by the neighbourhood information. This step further enhances the model’s performance by aligning classifiers with the learned class distributions. The proposed GeNGuide (Generalized Neighbour Guidance) framework works seamlessly without any modifications for several scenarios of class-incremental learning, namely (i) when the class-labels have varying amount of noise, including the noise-less case; (ii) when, in addition to the presence of noisy labels, the data is imbalanced, which makes the problem even more challenging. To the best of our knowledge, this is the first work which takes a step towards building generalized models which achieves state-of- the-art performance on several challenging class-incremental learning protocols, thereby justifying its effectiveness.
Contact: kumaravnish.456@gmail.com
This repository is tested in an Anaconda environment. To reproduce the results, create your environment as follows:
- Clone the repository:
git clone https://github.com/avnCode/GeNGuide.git
cd GeNGuide
- Create conda environment
conda create -y -n GeNGuide python=3.10
conda activate GeNGuide
- Install dependencies
conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.3 -c pytorch
pip install tqdm==4.66.4
pip install timm==0.5.4
pip install numpy==1.24.3
pip install scipy==1.11.1
pip install quadprog==0.1.11
pip install POT==0.9.1
pip install scikit-learn==1.3.2
pip install PyYAML
-
--config: Specify the path to the configuration file. -
--pretrained: Choose the pre-trained model to be used. The current options available are 'imagenet' or 'moco'. -
--noise_type: Define the type of noise to be applied.- For 'cifar10' dataset, you can use 'symmetric' or 'asymmetric'.
- For 'cifar100' dataset, 'random' or 'superclass' can be used.
python main.py --config=exps/genguide_cifar.json --noise=0.4 --noise_type='random' --dataset='cifar100' --pretrained='imagenet' Please download pre-trained ViT-Base models from MoCo v3 and ImaegNet-21K and then put or link the pre-trained models to GeNGuide/pretrained