English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In supervised continual learning, a deep neural network (DNN) is updated withan ever-growing data stream. Unlike the offline setting where data is shuffled,we cannot make any distributional assumptions about the data stream. Ideally,only one pass through the dataset is needed for computational efficiency.However, existing methods are inadequate and make many assumptions that cannotbe made for real-world applications, while simultaneously failing to improvecomputational efficiency. In this paper, we propose a novel continual learningmethod, SIESTA based on wake/sleep framework for training, which is wellaligned to the needs of on-device learning. The major goal of SIESTA is toadvance compute efficient continual learning so that DNNs can be updatedefficiently using far less time and energy. The principal innovations of SIESTAare: 1) rapid online updates using a rehearsal-free, backpropagation-free, anddata-driven network update rule during its wake phase, and 2) expedited memoryconsolidation using a compute-restricted rehearsal policy during its sleepphase. For memory efficiency, SIESTA adapts latent rehearsal using memoryindexing from REMIND. Compared to REMIND and prior arts, SIESTA is far morecomputationally efficient, enabling continual learning on ImageNet-1K in under2 hours on a single GPU; moreover, in the augmentation-free setting it matchesthe performance of the offline learner, a milestone critical to drivingadoption of continual learning in real-world applications.

SIESTA: Efficient Online Continual Learning with Sleep