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

Grouping has been commonly used in deep metric learning for computing diversefeatures. However, current methods are prone to overfitting and lackinterpretability. In this work, we propose an improved and interpretablegrouping method to be integrated flexibly with any metric learning framework.Our method is based on the attention mechanism with a learnable query for eachgroup. The query is fully trainable and can capture group-specific informationwhen combined with the diversity loss. An appealing property of our method isthat it naturally lends itself interpretability. The attention scores betweenthe learnable query and each spatial position can be interpreted as theimportance of that position. We formally show that our proposed grouping methodis invariant to spatial permutations of features. When used as a module inconvolutional neural networks, our method leads to translational invariance. Weconduct comprehensive experiments to evaluate our method. Our quantitativeresults indicate that the proposed method outperforms prior methodsconsistently and significantly across different datasets, evaluation metrics,base models, and loss functions. For the first time to the best of ourknowledge, our interpretation results clearly demonstrate that the proposedmethod enables the learning of distinct and diverse features across groups. Thecode is available on https://github.com/XinyiXuXD/DGML-master.

Towards Improved and Interpretable Deep Metric Learning via Attentive Grouping