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

Open-vocabulary models are a promising new paradigm for image classification.Unlike traditional classification models, open-vocabulary models classify amongany arbitrary set of categories specified with natural language duringinference. This natural language, called "prompts", typically consists of a setof hand-written templates (e.g., "a photo of a {}") which are completed witheach of the category names. This work introduces a simple method to generatehigher accuracy prompts, without relying on any explicit knowledge of the taskdomain and with far fewer hand-constructed sentences. To achieve this, wecombine open-vocabulary models with large language models (LLMs) to createCustomized Prompts via Language models (CuPL, pronounced "couple"). Inparticular, we leverage the knowledge contained in LLMs in order to generatemany descriptive sentences that contain important discriminatingcharacteristics of the image categories. This allows the model to place agreater importance on these regions in the image when making predictions. Wefind that this straightforward and general approach improves accuracy on arange of zero-shot image classification benchmarks, including over onepercentage point gain on ImageNet. Finally, this simple baseline requires noadditional training and remains completely zero-shot. Code available athttps://github.com/sarahpratt/CuPL.

What does a platypus look like? Generating customized prompts for zero-shot image classification