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

Recent breakthroughs of Neural Architecture Search (NAS) extend the field'sresearch scope towards a broader range of vision tasks and more diversifiedsearch spaces. While existing NAS methods mostly design architectures on asingle task, algorithms that look beyond single-task search are surging topursue a more efficient and universal solution across various tasks. Many ofthem leverage transfer learning and seek to preserve, reuse, and refine networkdesign knowledge to achieve higher efficiency in future tasks. However, theenormous computational cost and experiment complexity of cross-task NAS areimposing barriers for valuable research in this direction. Existing NASbenchmarks all focus on one type of vision task, i.e., classification. In thiswork, we propose TransNAS-Bench-101, a benchmark dataset containing networkperformance across seven tasks, covering classification, regression,pixel-level prediction, and self-supervised tasks. This diversity providesopportunities to transfer NAS methods among tasks and allows for more complextransfer schemes to evolve. We explore two fundamentally different types ofsearch space: cell-level search space and macro-level search space. With 7,352backbones evaluated on seven tasks, 51,464 trained models with detailedtraining information are provided. With TransNAS-Bench-101, we hope toencourage the advent of exceptional NAS algorithms that raise cross-task searchefficiency and generalizability to the next level. Our dataset file will beavailable at Mindspore, VEGA.

TransNAS-Bench-101: Improving Transferability and Generalizability of Cross-Task Neural Architecture Search