Shengbang Tong, Xili Dai, Ziyang Wu, Mingyang Li, Brent Yi, Yi Ma

This repository contains the implementation for the paper "Incremental Learning of Structured Memory via Closed-Loop Transcription". This work proposes a minimal computational model for learning structured memories of multiple object classes in an incremental setting. Our approach is based on establishing a closed-loop transcription between the classes and a corresponding set of subspaces, known as a linear discriminative representation, in a low-dimensional feature space. Network parameters are optimized simultaneously without architectural manipulations, by solving a constrained minimax game between the encoding and decoding maps over a single rate reduction-based objective. Experimental results show that our method can effectively alleviate catastrophic forgetting, achieving significantly better performance than prior work of generative replay on MNIST, CIFAR-10, and ImageNet-50, despite requiring fewer resources.

Current code implementation supports MNIST and cifar10. We will update more datasets in the near future~

To get started with the i-CTRL implementation, follow these instructions:

git clone https://github.com/tsb0601/i-CTRL.git
cd i-CTRL

pip install -r requirements.txt

The model and training configurations are as follows:

• PCACOMP: Number of principal components for PCA in nearsub evaluation (default = 15)
• SAMPLE_N: Number of sample classes (default = 12)
• SAMPLE_K: Number of samples per class (default = 40)
• LAMBD: Lambda parameter for reviewing (default = 10)
• LRG: Learning rate for generator (default = 0.0001)
• LRD: Learning rate for discriminator (default = 0.0001)
• EPOCHS: Training Epochs per Incremental Task (default = 100)

If you want to train MNIST, please use:

python main.py --cfg experiments/cifar10.yaml

If you want to train CIFAR-10, please use:

python main.py --cfg experiments/mnist.yaml

This repo is inspired by MCR2, EMP-SSL and CTRL repo.

If you find this repository useful, please consider giving a star ⭐ and citation:

@article{tong2022incremental,
  title={Incremental learning of structured memory via closed-loop transcription},
  author={Tong, Shengbang and Dai, Xili and Wu, Ziyang and Li, Mingyang and Yi, Brent and Ma, Yi},
  journal={arXiv preprint arXiv:2202.05411},
  year={2022}
}