Continual Classification Learning Using Generative Models

Lavda, Frantzeska; Ramapuram, Jason; Gregorova, Magda; Kalousis, Alexandros

doi:10.48550/arxiv.1810.10612

Cited by 11 publications

(13 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Rehearsal iCaRL [18] ER [44] SER [45] TEM [46] Pseudo Rehearsal DGR [14] PR [47] CCLUGM [48] LGM [49] Constrained GEM [50] A-GEM [8] GSS [43] Regularization-based methods…”

Section: Replay-based Methodsmentioning

confidence: 99%

A continual learning survey: Defying forgetting in classification tasks

De Lange,

Aljundi,

Masana

et al. 2019

Preprint

139

View full text Add to dashboard Cite

Artificial neural networks thrive in solving the classification problem for a particular rigid task, where the network resembles a static entity of knowledge, acquired through generalized learning behaviour from a distinct training phase. However, endeavours to extend this knowledge without targeting the original task usually result in a catastrophic forgetting of this task. Continual learning shifts this paradigm towards a network that can continually accumulate knowledge over different tasks without the need for retraining from scratch, with methods in particular aiming to alleviate forgetting. We focus on task-incremental classification, where tasks arrive in a batch-like fashion, and are delineated by clear boundaries. Our main contributions concern 1) a taxonomy and extensive overview of the state-of-the-art, 2) a novel framework to continually determine stability-plasticity trade-off of the continual learner, 3) a comprehensive experimental comparison of 10 state-of-the-art continual learning methods and 4 baselines.We empirically scrutinize which method performs best, both on balanced Tiny Imagenet and a large-scale unbalanced iNaturalist datasets. We study the influence of model capacity, weight decay and dropout regularization, and the order in which the tasks are presented, and qualitatively compare methods in terms of required memory, computation time and storage. We make code publicly available upon acceptance of this paper.

show abstract

“…Rehearsal iCaRL [18] ER [44] SER [45] TEM [46] Pseudo Rehearsal DGR [14] PR [47] CCLUGM [48] LGM [49] Constrained GEM [50] A-GEM [8] GSS [43] Regularization-based methods…”

Section: Replay-based Methodsmentioning

confidence: 99%

A continual learning survey: Defying forgetting in classification tasks

De Lange,

Aljundi,

Masana

et al. 2019

Preprint

139

View full text Add to dashboard Cite

show abstract

“…It is also worth mentioning that other works [28], [29] exploit generative models in place of the memory buffer. On the one hand, this favorably allows sampling data-points from the distributions underlying the old tasks.…”

Section: Related Workmentioning

confidence: 99%

Rethinking Experience Replay: a Bag of Tricks for Continual Learning

Buzzega

Boschini²,

Porrello

et al. 2021

2020 25th International Conference on Pattern Recognition (ICPR)

126

125

View full text Add to dashboard Cite

In Continual Learning, a Neural Network is trained on a stream of data whose distribution shifts over time. Under these assumptions, it is especially challenging to improve on classes appearing later in the stream while remaining accurate on previous ones. This is due to the infamous problem of catastrophic forgetting, which causes a quick performance degradation when the classifier focuses on learning new categories. Recent literature proposed various approaches to tackle this issue, often resorting to very sophisticated techniques. In this work, we show that naive rehearsal can be patched to achieve similar performance. We point out some shortcomings that restrain Experience Replay (ER) and propose five tricks to mitigate them. Experiments show that ER, thus enhanced, displays an accuracy gain of 51.2 and 26.9 percentage points on the CIFAR-10 and CIFAR-100 datasets respectively (memory buffer size 1000). As a result, it surpasses current state-of-the-art rehearsal-based methods.

show abstract

“…Loss of prior knowledge is therefore avoided by periodic refreshing of the classifier's memory with data that represent such prior knowledge. Deep generative replay (DGR) [10], deep generative replay with distillation [5], [11], and continual classification using generative models [12] are examples of generative replay-based approaches.…”

Section: A Continual Learningmentioning

confidence: 99%

Adversarial Targeted Forgetting in Regularization and Generative Based Continual Learning Models

Umer¹,

Polikar²

2021

Preprint

View full text Add to dashboard Cite

Continual (or "incremental") learning approaches are employed when additional knowledge or tasks need to be learned from subsequent batches or from streaming data. However these approaches are typically adversary agnostic, i.e., they do not consider the possibility of a malicious attack. In our prior work, we explored the vulnerabilities of Elastic Weight Consolidation (EWC) to the perceptible misinformation. We now explore the vulnerabilities of other regularization-based as well as generative replay-based continual learning algorithms, and also extend the attack to imperceptible misinformation. We show that an intelligent adversary can take advantage of a continual learning algorithm's capabilities of retaining existing knowledge over time, and force it to learn and retain deliberately introduced misinformation. To demonstrate this vulnerability, we inject backdoor attack samples into the training data. These attack samples constitute the misinformation, allowing the attacker to capture control of the model at test time. We evaluate the extent of this vulnerability on both rotated and split benchmark variants of the MNIST dataset under two important domain and class incremental learning scenarios. We show that the adversary can create a "false memory" about any task by inserting carefullydesigned backdoor samples to the test instances of that task thereby controlling the amount of forgetting of any task of its choosing. Perhaps most importantly, we show this vulnerability to be very acute and damaging: the model memory can be easily compromised with the addition of backdoor samples into as little as 1% of the training data, even when the misinformation is imperceptible to human eye.

show abstract

Continual Classification Learning Using Generative Models

Cited by 11 publications

References 6 publications

A continual learning survey: Defying forgetting in classification tasks

A continual learning survey: Defying forgetting in classification tasks

Rethinking Experience Replay: a Bag of Tricks for Continual Learning

Adversarial Targeted Forgetting in Regularization and Generative Based Continual Learning Models

Contact Info

Product

Resources

About