Dataset Condensation Atlas

Application · Meta-learning through unrolled training

CCMCL

Condensed Composite Memory Continual Learning

Felix Wiewel, Bin Yang

IJCNN 2021 · first public 2021-02-19 · arXiv 2102.09890

paper ↗code ↗catalogued✓ abstract read

In one paragraph

For rehearsal-based continual learning, learns a small set of shared components combined into per-class synthetic exemplars rather than learning each replay image directly, increasing memory efficiency of the stored rehearsal buffer. Reports improved performance over prior rehearsal baselines, especially at very small buffer sizes.

Where it sits

Design choices

What is storedfactorized-bases
Abstract (verbatim from arXiv)

Deep Neural Networks (DNNs) suffer from a rapid decrease in performance when trained on a sequence of tasks where only data of the most recent task is available. This phenomenon, known as catastrophic forgetting, prevents DNNs from accumulating knowledge over time. Overcoming catastrophic forgetting and enabling continual learning is of great interest since it would enable the application of DNNs in settings where unrestricted access to all the training data at any time is not always possible, e.g. due to storage limitations or legal issues. While many recently proposed methods for continual learning use some training examples for rehearsal, their performance strongly depends on the number of stored examples. In order to improve performance of rehearsal for continual learning, especially for a small number of stored examples, we propose a novel way of learning a small set of synthetic examples which capture the essence of a complete dataset. Instead of directly learning these synthetic examples, we learn a weighted combination of shared components for each example that enables a significant increase in memory efficiency. We demonstrate the performance of our method on commonly used datasets and compare it to recently proposed related methods and baselines.

BibTeX (generated; prefer the venue's official entry)
@article{wiewel2021condensed,
  title   = {Condensed Composite Memory Continual Learning},
  author  = {Felix Wiewel and Bin Yang},
  journal = {IJCNN 2021},
  year    = {2021}
}

Nearby in Meta-learning through unrolled training

2024-10

Offline Behavior Distillation

Shiye Lei, Sen Zhang, Dacheng Tao · NeurIPS 2024notableOther datapaper ↗code ↗

2023-11

RaT-BPTT — Embarassingly Simple Dataset Distillation

Yunzhen Feng, Ramakrishna Vedantam, Julia Kempe · ICLR 2024notablepaper ↗code ↗

2019-12

GTN — Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data

Felipe Petroski Such, Aditya Rawal, Joel Lehman et al. · ICML 2020notablepaper ↗code ↗

2018-11

DD — Dataset Distillation

Tongzhou Wang, Jun-Yan Zhu, Antonio Torralba et al. · arXiv 2018landmarkpaper ↗code ↗