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

Recently, self-supervised metric learning has raised attention for thepotential to learn a generic distance function. It overcomes the limitations ofconventional supervised one, e.g., scalability and label biases. Despiteprogress in this domain, current benchmarks, incorporating a narrow scope ofclasses, stop the nuanced evaluation of semantic representations. To bridgethis gap, we introduce a large-scale benchmark with diversity and granularityof classes, Statistical Metric Learning Benchmark (SMLB) built uponImageNet-21K and WordNet. SMLB is designed to rigorously evaluate thediscriminative discernment and generalizability across more than 14M images,20K classes, and 16K taxonomic nodes. Alongside, we propose novel evaluationmetrics -- `overlap' for separability and `aSTD' for consistency -- to measuredistance statistical information, which are efficient and robust to the changeof class number. Our benchmark offers a novel perspective of evaluating thequality of representations beyond accuracy. Our findings reveal the limitationsof supervised learning and the class bias inherent in SSL models, offeringinsights into potential areas for future model enhancement.

Beyond Accuracy: Statistical Measures and Benchmark for Evaluation of Representation from Self-Supervised Learning