Sample Condensation in Online Continual Learning
This addresses the problem of memory efficiency in continual learning for AI systems, but it is incremental as it builds on existing replay strategies.
The paper tackles catastrophic forgetting in online continual learning by proposing OLCGM, a replay-based strategy that uses sample condensation to compress old samples in memory, improving final accuracy compared to state-of-the-art methods when memory is limited relative to data complexity.
Online Continual learning is a challenging learning scenario where the model must learn from a non-stationary stream of data where each sample is seen only once. The main challenge is to incrementally learn while avoiding catastrophic forgetting, namely the problem of forgetting previously acquired knowledge while learning from new data. A popular solution in these scenario is to use a small memory to retain old data and rehearse them over time. Unfortunately, due to the limited memory size, the quality of the memory will deteriorate over time. In this paper we propose OLCGM, a novel replay-based continual learning strategy that uses knowledge condensation techniques to continuously compress the memory and achieve a better use of its limited size. The sample condensation step compresses old samples, instead of removing them like other replay strategies. As a result, the experiments show that, whenever the memory budget is limited compared to the complexity of the data, OLCGM improves the final accuracy compared to state-of-the-art replay strategies.