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

Learning the embedding space, where semantically similar objects are locatedclose together and dissimilar objects far apart, is a cornerstone of manycomputer vision applications. Existing approaches usually learn a single metricin the embedding space for all available data points, which may have a verycomplex non-uniform distribution with different notions of similarity betweenobjects, e.g. appearance, shape, color or semantic meaning. Approaches forlearning a single distance metric often struggle to encode all different typesof relationships and do not generalize well. In this work, we propose a noveleasy-to-implement divide and conquer approach for deep metric learning, whichsignificantly improves the state-of-the-art performance of metric learning. Ourapproach utilizes the embedding space more efficiently by jointly splitting theembedding space and data into $K$ smaller sub-problems. It divides both, thedata and the embedding space into $K$ subsets and learns $K$ separate distancemetrics in the non-overlapping subspaces of the embedding space, defined bygroups of neurons in the embedding layer of the neural network. The proposedapproach increases the convergence speed and improves generalization since thecomplexity of each sub-problem is reduced compared to the original one. We showthat our approach outperforms the state-of-the-art by a large margin inretrieval, clustering and re-identification tasks on CUB200-2011, CARS196,Stanford Online Products, In-shop Clothes and PKU VehicleID datasets.

Divide and Conquer the Embedding Space for Metric Learning