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

Existing object recognition models have been shown to lack robustness indiverse geographical scenarios due to significant domain shifts in design andcontext. Class representations need to be adapted to more accurately reflect anobject concept under these shifts. In the absence of training data from targetgeographies, we hypothesize that geography-specific descriptive knowledge ofobject categories can be leveraged to enhance robustness. For this purpose, weexplore the feasibility of probing a large-language model forgeography-specific object knowledge, and we investigate integrating knowledgein zero-shot and learnable soft prompting with the CLIP vision-language model.In particular, we propose a geography knowledge regularization method to ensurethat soft prompts trained on a source set of geographies generalize to anunseen target set of geographies. Our gains on DollarStreet when generalizingfrom a model trained only on data from Europe are as large as +2.8 on countriesfrom Africa, and +4.6 on the hardest classes. We further show competitiveperformance vs. few-shot target training, and provide insights into howdescriptive knowledge captures geographical differences.

Incorporating Geo-Diverse Knowledge into Prompting for Increased Geographical Robustness in Object Recognition