论文阅读（略读）

Paper:Class-incremental learning: survey and performance evaluation

1. Abstract : 对现有12增量学习进行广泛评估；
2. Introduction: 早期的工作都是任务增量学习（task-incremental learning (task-IL)); 近期工作多为类增量学习 class-incremental learning (class- IL); 区别：在推理时是否可以访问任务ID;
   2.1 regularization-based solutions: 尽量减少学习新任务对前期任务的重要权重的影响； Exemplar-based:储存有限的范例; task-recency bias
3. Related Work:
   3.1 Task-incremental learning: Mask-based methods :通过对每个参数或者每层表征应用掩码来消除遗忘； Piggyback:学习mask在网络权重上； PackNet:学习权重然后prunes； PathNET使用进化策略通过权值选择路由；
   3.2 Online: Online methods are based on streaming frameworks where learners are allowed to observe each ex- ample only once instead of iterating over a set of examples in a training session.
   3.3 Variational continual learning: Variational continual learning is based on the Bayesian inference framework
   3.4 Pseudo-rehearsal methods. In order to avoid storing ex- emplars and privacy issues inherent in exemplar rehearsal,some methods learn to generate examples from previous tasks（生成范例）
4. CLASS-INCREMENTAL LEARNING
   4.1 Challenges of class-incremental learning
   4.1.1 权重漂移，Activation drift，Inter-task confusion，Task-recency bias
5. Approaches:
   5.1.1 基于Weight regularization
   RWARK(EWC, PATHIT), MAS
   5.1.2 基于 DATA regularization
   LWF
   5.1.3 Rehearsal approaches
   (1) Memory Types
   (2) Sampling strategies
   (3) Task balancing
   (4) Combining
   5.1.4 Bias-correction approaches
   EEIL(Balanced training) BIC(bias correction) LUCIR(cos,margin rank ),IL2M(rectifies based on the saved certainty statistics of predictions of classes from previous tasks)
6. relations
   7 experimental SETUP
   7.1 DATASET
   7.2 metrics
   Average accuracy, forgetting, intransigence,confusion matrix
   7.3 Experiment