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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Large language models (LLMs)-based image captioning has the capability ofdescribing objects not explicitly observed in training data; yet novel objectsoccur frequently, necessitating the requirement of sustaining up-to-date objectknowledge for open-world comprehension. Instead of relying on large amounts ofdata and scaling up network parameters, we introduce a highly effectiveretrieval-augmented image captioning method that prompts LLMs with object namesretrieved from External Visual--name memory (EVCap). We build ever-changingobject knowledge memory using objects' visuals and names, enabling us to (i)update the memory at a minimal cost and (ii) effortlessly augment LLMs withretrieved object names utilizing a lightweight and fast-to-train model. Ourmodel, which was trained only on the COCO dataset, can be adapted to out-domaindata without additional fine-tuning or retraining. Our comprehensiveexperiments conducted on various benchmarks and synthetic commonsense-violatingdata demonstrate that EVCap, comprising solely 3.97M trainable parameters,exhibits superior performance compared to other methods of equivalent modelsize scale. Notably, it achieves competitive performance against specialistSOTAs with an enormous number of parameters. Our code is available athttps://jiaxuan-li.github.io/EVCap.

EVCap: Retrieval-Augmented Image Captioning with External Visual-Name Memory for Open-World Comprehension