A wholistic view of continual learning with deep neural networks: Forgotten lessons and the bridge to active and open world learning.

Current deep learning methods are regarded as favorable if they empirically perform well on dedicated test sets. This mentality is seamlessly reflected in the resurfacing area of continual learning, where consecutively arriving data is investigated. The core challenge is framed as protecting previously acquired representations from being catastrophically forgotten. However, comparison of individual methods is nevertheless performed in isolation from the real world by monitoring accumulated benchmark test set performance. The closed world assumption remains predominant, i.e. models are evaluated on data that is guaranteed to originate from the same distribution as used for training. This poses a massive challenge as neural networks are well known to provide overconfident false predictions on unknown and corrupted instances. In this work we critically survey the literature and argue that notable lessons from open set recognition, identifying unknown examples outside of the observed set, and the adjacent field of active learning, querying data to maximize the expected performance gain, are frequently overlooked in the deep learning era. Hence, we propose a consolidated view to bridge continual learning, active learning and open set recognition in deep neural networks. Finally, the established synergies are supported empirically, showing joint improvement in alleviating catastrophic forgetting, querying data, selecting task orders, while exhibiting robust open world application.

Emotions are associated with the genesis of visually induced motion sickness in virtual reality.

Visually induced motion sickness (VIMS) is a well-known side effect of virtual reality (VR) immersion, with symptoms including nausea, disorientation, and oculomotor discomfort. Previous studies have shown that pleasant music, odor, and taste can mitigate VIMS symptomatology, but the mechanism by which this occurs remains unclear. We predicted that positive emotions influence the VIMS-reducing effects. To investigate this, we conducted an experimental study with 68 subjects divided into two groups. The groups were exposed to either positive or neutral emotions before and during the VIMS-provoking stimulus. Otherwise, they performed exactly the same task of estimating the time-to-contact while confronted with a VIMS-provoking moving starfield stimulation. Emotions were induced by means of pre-tested videos and with International Affective Picture System (IAPS) images embedded in the starfield simulation. We monitored emotion induction before, during, and after the simulation, using the Self-Assessment Manikin (SAM) valence and arousal scales. VIMS was assessed before and after exposure using the Simulator Sickness Questionnaire (SSQ) and during simulation using the Fast Motion Sickness Scale (FMS) and FMS-D for dizziness symptoms. VIMS symptomatology did not differ between groups, but valence and arousal were correlated with perceived VIMS symptoms. For instance, reported positive valence prior to VR exposure was found to be related to milder VIMS symptoms and, conversely, experienced symptoms during simulation were negatively related to subjects' valence. This study sheds light on the complex and potentially bidirectional relationship of VIMS and emotions and provides starting points for further research on the use of positive emotions to prevent VIMS.

Return of the normal distribution: Flexible deep continual learning with variational auto-encoders.

Learning continually from sequentially arriving data has been a long standing challenge in machine learning. An emergent body of deep learning literature suggests various solutions, through introduction of significant simplifications to the problem statement. As a consequence of a growing focus on particular tasks and their respective benchmark assumptions, these efforts are thus becoming increasingly tailored to specific settings. Whereas approaches that leverage Variational Bayesian techniques seem to provide a more general perspective of key continual learning mechanisms, they however entail their own caveats. Inspired by prior theoretical work on solving the prevalent mismatch between prior and aggregate posterior in deep generative models, we return to a generic variational auto-encoder based formulation and investigate its utility for continual learning. Specifically, we propose to adapt a two-stage training framework towards a context conditioned variant for continual learning, where we then formulate mechanisms to alleviate catastrophic forgetting through choices of generative rehearsal or well-motivated extraction of data exemplar subsets. Although the proposed generic two-stage variational auto-encoder is not tailored towards a particular task and allows for flexible amounts of supervision, we empirically demonstrate it to surpass task-tailored methods in both supervised classification, as well as unsupervised representation learning.

Unified Probabilistic Deep Continual Learning through Generative Replay and Open Set Recognition.

Modern deep neural networks are well known to be brittle in the face of unknown data instances and recognition of the latter remains a challenge. Although it is inevitable for continual-learning systems to encounter such unseen concepts, the corresponding literature appears to nonetheless focus primarily on alleviating catastrophic interference with learned representations. In this work, we introduce a probabilistic approach that connects these perspectives based on variational inference in a single deep autoencoder model. Specifically, we propose to bound the approximate posterior by fitting regions of high density on the basis of correctly classified data points. These bounds are shown to serve a dual purpose: unseen unknown out-of-distribution data can be distinguished from already trained known tasks towards robust application. Simultaneously, to retain already acquired knowledge, a generative replay process can be narrowed to strictly in-distribution samples, in order to significantly alleviate catastrophic interference.