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 this paper, we introduce a novel approach for systematically solvingdataset condensation problem in an efficient manner by exploiting theregularity in a given dataset. Instead of condensing the dataset directly inthe original input space, we assume a generative process of the dataset with aset of learnable codes defined in a compact latent space followed by a set oftiny decoders which maps them differently to the original input space. Bycombining different codes and decoders interchangeably, we can dramaticallyincrease the number of synthetic examples with essentially the same parametercount, because the latent space is much lower dimensional and since we canassume as many decoders as necessary to capture different styles represented inthe dataset with negligible cost. Such knowledge factorization allows efficientsharing of information between synthetic examples in a systematic way,providing far better trade-off between compression ratio and quality of thegenerated examples. We experimentally show that our method achieves newstate-of-the-art records by significant margins on various benchmark datasetssuch as SVHN, CIFAR10, CIFAR100, and TinyImageNet.

Dataset Condensation with Latent Space Knowledge Factorization and Sharing