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

Finetuning a pretrained vision model (PVM) is a common technique for learningdownstream vision tasks. The conventional finetuning process with the randomlysampled data points results in diminished training efficiency. To address thisdrawback, we propose a novel approach, VLM-empowered Collaborative ActiveFinetuning (VeCAF). VeCAF optimizes a parametric data selection model byincorporating the training objective of the model being tuned. Effectively,this guides the PVM towards the performance goal with improved data andcomputational efficiency. As vision-language models (VLMs) have achievedsignificant advancements by establishing a robust connection between image andlanguage domains, we exploit the inherent semantic richness of the textembedding space and utilize text embedding of pretrained VLM models to augmentPVM image features for better data selection and finetuning. Furthermore, theflexibility of text-domain augmentation gives VeCAF a unique ability to handleout-of-distribution scenarios without external augmented data. Extensiveexperiments show the leading performance and high efficiency of VeCAF that issuperior to baselines in both in-distribution and out-of-distribution imageclassification tasks. On ImageNet, VeCAF needs up to 3.3x less training batchesto reach the target performance compared to full finetuning and achieves 2.8%accuracy improvement over SOTA methods with the same number of batches.

VeCAF: VLM-empowered Collaborative Active Finetuning with Training Objective Awareness