用新数据复活旧类：无样本持续学习
Resurrecting Old Classes with New Data for Exemplar-Free Continual Learning
AI总结：
针对无样本持续学习中的特征漂移问题，提出用对抗扰动样本估计原型移动并补偿，在多个基准上超越现有方法。
摘要：
持续学习方法已知会遭受灾难性遗忘，这一现象对于不存储先前任务样本的方法尤其难以应对。因此，为了减少特征提取器中潜在的漂移，现有的无样本方法通常在第一个任务显著大于后续任务的设置下进行评估。在从较小的第一个任务开始的更具挑战性的设置中，它们的性能急剧下降。为了解决无样本方法的特征漂移估计问题，我们提出对抗性地扰动当前样本，使其嵌入在旧模型嵌入空间中接近旧类原型。然后，我们使用扰动图像估计从旧模型到新模型的嵌入空间中的漂移，并相应地补偿原型。我们利用了对抗样本在持续学习设置中从旧特征空间到新特征空间的可转移性。这些图像的生成简单且计算成本低。我们的实验表明，所提出的方法能更好地跟踪嵌入空间中原型的移动，并在多个标准持续学习基准以及细粒度数据集上优于现有方法。代码可在 https://github.com/dipamgoswami/ADC 获取。
英文摘要：
Continual learning methods are known to suffer from catastrophic forgetting, a phenomenon that is particularly hard to counter for methods that do not store exemplars of previous tasks. Therefore, to reduce potential drift in the feature extractor, existing exemplar-free methods are typically evaluated in settings where the first task is significantly larger than subsequent tasks. Their performance drops drastically in more challenging settings starting with a smaller first task. To address this problem of feature drift estimation for exemplar-free methods, we propose to adversarially perturb the current samples such that their embeddings are close to the old class prototypes in the old model embedding space. We then estimate the drift in the embedding space from the old to the new model using the perturbed images and compensate the prototypes accordingly. We exploit the fact that adversarial samples are transferable from the old to the new feature space in a continual learning setting. The generation of these images is simple and computationally cheap. We demonstrate in our experiments that the proposed approach better tracks the movement of prototypes in embedding space and outperforms existing methods on several standard continual learning benchmarks as well as on fine-grained datasets. Code is available at https://github.com/dipamgoswami/ADC.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
