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 Zero-shot learning (ZSL), we classify unseen categories using textualdescriptions about their expected appearance when observed (class embeddings)and a disjoint pool of seen classes, for which annotated visual data areaccessible. We tackle ZSL by casting a "vanilla" convolutional neural network(e.g. AlexNet, ResNet-101, DenseNet-201 or DarkNet-53) into a zero-shotlearner. We do so by crafting the softmax classifier: we freeze its weightsusing fixed seen classification rules, either semantic (seen class embeddings)or visual (seen class prototypes). Then, we learn a data-driven andZSL-tailored feature representation on seen classes only to match these fixedclassification rules. Given that the latter seamlessly generalize towardsunseen classes, while requiring not actual unseen data to be computed, we canperform ZSL inference by augmenting the pool of classification rules at testtime while keeping the very same representation we learnt: nowhere re-trainingor fine-tuning on unseen data is performed. The combination of semantic andvisual crafting (by simply averaging softmax scores) improves priorstate-of-the-art methods in benchmark datasets for standard, inductive ZSL.After rebalancing predictions to better handle the joint inference over seenand unseen classes, we outperform prior generalized, inductive ZSL methods aswell. Also, we gain interpretability at no additional cost, by using neuralattention methods (e.g., grad-CAM) as they are. Code will be made publiclyavailable.

Classifier Crafting: Turn Your ConvNet into a Zero-Shot Learner!